Guitar Barry. Stuttering An Integrated Approach to Its Nature and Treatment

SECTION 1 Nature of Stuttering
1 Introduction to Stuttering
Overview of the Disorder
The Human Face of Stuttering
Basic Facts about Stuttering and their Implications for the Nature of Stuttering
2 Constitutional Factors in Stuttering
Biological Background
Sensory and Sensory-Motor Factors
Language Factors
Emotional Factors
A Lay Description of Constitutional Factors
3 Research Findings about Constitutional Factors in Stuttering
Hereditary Factors
Congenital and Early Childhood Trauma Studies
Brain Structure and Function
Sensory and Sensory-Motor Studies
Sensory-Motor Control
Language Factors
Emotional Factors
4 Developmental, Environmental, and Learning Factors in Stuttering
Developmental Factors
Environmental Factors
Learning Factors
5 Research Findings on Developmental, Environmental, and LearningFactors in Stuttering
Physical and Motor Skill Development
Speech and Language Development
Cognitive Development
Environmental Factors
6 Theories about Stuttering
Theoretical Perspectives about Constitutional Factors in Stuttering
Theoretical Perspectives on Developmental and Environmental Factors An Integration of Perspectives on Stuttering
7 Normal Disfluency and the Development of Stuttering
Normal Disfluency
Younger Preschool Children: Borderline Stuttering
Older Preschool Children: Beginning Stuttering
School-Age Children: Intermediate Stuttering
Older Teens and Adults: Advanced Stuttering
SECTION 2 Assessment and Treatment of Stuttering
8 Preliminaries to Assessment
The Client’s Needs
The Client’s Right to Privacy Cultural Considerations
The Clinician’s Expertise
Assessment of Stuttering Behavior
Assessing Speech Naturalness
Assessing Speaking and Reading Rate
Assessing Feelings and Attitudes
Continuing Assessment
9 Assessment and Diagnosis
Preschool Child
School-Age Child
Adolescents and Adults
10 Preliminaries to Treatment
Clinician’s Attributes
Clinician’s Beliefs
Treatment Goals
Therapy Procedures
11 Treatment of Stuttering in Younger Preschool Children: Borderline
An Integrated Approach Other Clinicians
12 Treatment of Stuttering in Older Preschool Children: Beginning
An Integrated Approach
Another Clinician
Treatment of Concomitant Speech and Language Problems
13 Treatment of School-Age Children: Intermediate Stuttering
An Integrated Approach
Other Clinicians
14 Treatment of Stuttering in Adolescents and Adults: Advanced
An Integrated Approach
Other Approaches
15 Related Disorders of Fluency
Neurogenic Acquired Stuttering
Psychogenic Acquired Stuttering Cluttering References

Introduction to Stuttering

The Words We Use
People Who Stutter
Overview of the Disorder Do All Cultures Have Stuttering?
What Causes People to Stutter?
Can Stuttering Be Cured?
General Description
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Functioning, Disability, and Health
The Human Face of Stuttering
Basic Facts about Stuttering and Their Implications for the Nature of Stuttering
Recovery without Treatment
Sex Ratio
Variability and Predictability of Stuttering
Anticipation, Consistency, and Adaptation
Language Factors
Fluency-Inducing Conditions
An Integration
After studying this chapter, readers should be able to:
• Explain why it is good practice to use the term “person who stutters” ratherthan “stutterer”
• Describe factors that may (a) predispose a child to stutter, (b) precipitate stuttering, and (c) make stuttering persistent
• Name and describe the core behaviors of stuttering
• Name and describe the two major categories of secondary stuttering behaviors
• Name and describe different feelings and attitudes that can accompany stuttering
• Describe the elements of the new International Classification of Functioning,Disability, and Health (ICF) system that are most relevant to stuttering
• Discuss the age range of stuttering onset and the types of onset, and explainwhy the onset of stuttering is often difficult to pinpoint
• Describe the meanings of the terms “prevalence” and “incidence,” and givecurrent best estimates of each
• Give an estimate of the number of children who recover without treatment, anddescribe factors that predict this recovery
• Give an estimate of the sex ratio in stuttering at onset and in the school-agepopulation
• Explain what is meant by “anticipation,” “consistency,” and “adaptation” instuttering
• Explain some relationships between stuttering and language, and suggest whatthey mean about the nature of the disorder
• Describe several conditions under which stuttering is usually reduced or absent,and suggest why this may be so
Disfluency: An interruption of speech–such as a repetition, hesitancy, or prolongation of sound–that may occur in both typically developing individuals and those who stutter
Normal disfluency: An interruption of speech in a typically developing individual
Fluency: The effortless flow of speech
Repetition: A sound, syllable, or single-syllable word that is repeated several times. The speaker is apparently “stuck” on that sound and continues
repeating it until the following sound can be produced
Prolongation: A stutter in which sound or air flow continues but movement of the articulators is stopped
Block: A stutter that is an inappropriate stoppage of the flow of air or voice and often the movement of articulators as well
Core behavior: The basic speech behavior of stuttering–repetition, prolongation, and block
Secondary behavior: A speaker’s reactions to his or her repetitions, prolongations, and blocks in an attempt to end them quickly or avoid them altogether. Such reactions may begin as random struggle but soon turn into well-learned patterns. Secondary behaviors can be divided into two broad classes: escape and avoidance behaviors
Escape behavior: A speaker’s attempts to terminate a stutter and finish the word. This occurs when the speaker is already in a moment of stuttering
Avoidance behavior: A speaker’s attempt to prevent stuttering when he or she anticipates stuttering on a word or in a situation. Word-based avoidances are commonly interjections of extra sounds, like “uh,” said before the word on which stuttering is expected
Attitude: A feeling that has become a pervasive part of a person’s beliefs
Heterogeneity: Differences among various types of a disorder
Developmental stuttering: A term used to denote the most common form of stuttering that develops during childhood (in contrast to stuttering that develops in response to a neurological event or trauma or emotional stress)
Prevalence: A term used to indicate how widespread a disorder is
Incidence: An index of how many people have stuttered at some time in their lives
Anticipation: An individual’s ability to predict on which words or sounds he or she will stutter
Consistency: The tendency for speakers to stutter on the same words when reading a passage several times
Adaptation: The tendency for speakers to stutter less and less (up to a point) when repeatedly reading a passage
No one is sure what causes stuttering, but it is an age-old problem that may have its origins in the way our brains evolved to produce speech and language. Its sudden appearance in some children is triggered when they try to talk using their justemerging speech and language skills. Its many variations and manifestations are determined by individual learning patterns, personality, and temperament. It also provides lessons about human nature: the variety of responses that stuttering provokes in cultures around the world is a reflection of the many ways in which humans deal with individual differences.
This description of stuttering makes it seem like a very complicated problem–one that will take a long time to learn about. It’s true that you could spend a lifetime and still not know everything there is to know about stuttering. But you don’t need to understand everything in order to help people who stutter. If you read this book critically and carefully, you will get a basic understanding of stuttering and a foundation for evaluating and treating people who stutter and their families. And once you start working with people who stutter, your understanding and ability can expand exponentially.
If you continue to work with stuttering, you will soon outgrow this book and begin to make your own discoveries. You will experience the satisfaction of helping children, adolescents, and adults regain an ability to communicate easily. Someday you may even write about your therapy procedures and your assessment of their effectiveness. Those of us who have spent many years engaged in stuttering research and treatment all began where you are right now, at the threshold of an exciting and rewarding profession.
The Words We Use
In any field–whether it’s law, medicine, or speech-language pathology–words may be used in specific ways. Definitions of many of the specialized terms used in our field are provided in the glossary at the back of this book. But some words and phrases deserve to be discussed at the beginning.
People Who Stutter
Until recently, it was common practice to refer to people who stutter as “stutterers.” In fact, some of us who stutter refer to ourselves as stutterers and feel some pride in this term. However, many people prefer not to be labeled “a stutterer” and want to be called “people who stutter.” They feel, and rightly so, that stuttering is only a small part of who they are.
Adults who stutter often say that changing the way they think of themselves–as people who happen to stutter but with many more important attributes–was one of the most significant things they did to break free of the bonds of stuttering. Such reports remind us that clients are far more than people who stutter. They are people, each with a galaxy of characteristics, one of which happens to be that they stutter. This way of thinking enables us to help not only our clients but also their families. Families learn to listen beyond the sounds of stuttering to the thoughts and feelings that their children are communicating. It helps everyone view disfluencies in perspective as only a small part of the whole child.
Some authors abbreviate “people who stutter” as “PWS.” Personally, I feel that substituting an acronym that highlights stuttering is not really different from using “stutterer,” so I won’t employ that as a euphemism. However, I know that the language in this book would grow stale and cumbersome if I were to use “person who stutters” over and over. So I often refer to the “adult,” “child,” or “adolescent you are working with,” and may sometimes use “stutterer.” That, too, I feel is acceptable when used occasionally. After all, a stutterer may be someone who is proud that he sometimes stutters but doesn’t let it get in the way of his life.
In our literature, “disfluency” is used to denote interruptions of speech that may be either normal or abnormal. That is, it can apply to pauses, repetitions, and other hesitancies that occur in the speech of persons with normal speech. It can also apply to moments of stuttering and is a handy term to use when describing the speech of young children whose diagnosis is unclear.
I’ll use “disfluency” interchangeably with “stutter” to make the writing more varied. When someone’s speech hesitancies are unequivocally not stuttering, I’ll use the term “normal disfluency.” I won’t use the older term for abnormal hesitations, “dysfluency” with a “y,” because it can easily be mistaken for “disfluency” when you see it on the page and because the two are indistinguishable when spoken.
This section previews the next few chapters on the nature of stuttering and gives me a chance to reveal my own slant on the disorder. I think this may be helpful for anyone, but especially for those readers who have not had a course in stuttering and who may, therefore, know few details of its nature.
Do All Cultures Have Stuttering?
Stuttering is found in all parts of the world and in all cultures and races. It is indiscriminate of occupation, intelligence, and income; it affects both sexes and people of all ages, from toddlers to the elderly. It is an old curse, and there is evidence that it was present in Chinese, Egyptian, and Mesopotamian cultures more than 40 centuries ago. Moses was said to have stuttered (Garfinkel, 1995) and to have used a trick typical of many of us who stutter–getting his brother to speak for him. I did something similar when I was asked to read a prayer aloud in Sunday school.
What Causes People to Stutter?
The cause of stuttering is still something of a mystery. Scientists have yet to discover what causes stuttering, but they have many clues. First, there is strong evidence that stuttering often has a genetic basis–that is, something is inherited that makes it more likely a child will stutter. This genetic “something” has to do with the way a child’s brain develops its neural pathways for speech and language. For example, the neural pathways for talking may have bottlenecks, dead ends, or other obstacles to the rapid flow of information. The pathways may also be vulnerable to disruption by other brain activity, such as emotions. Another clue about the nature of stuttering is that most stuttering begins in children between ages 2 and 5. Thus, the onset of stuttering occurs about the same time that many typical stresses of early childhood are occurring. One child may begin to stutter during a dramatic growth in vocabulary and syntax. Another’s stuttering may first appear when the family moves to a new home. Still another child may start soon after a baby brother or sister is born. Many different factors, acting singly or in combination, may precipitate the onset of stuttering in a child who has a neurophysiological predisposition, or inborn tendency, for stuttering.
Once stuttering starts, it may disappear within a few months, or it may get gradually worse. When it gets worse, learned reactions may be an important factor in its severity. Playmates at school or thoughtless adults may cause a child to become highly self-conscious about his stuttering. The child will quickly learn that by pushing hard, he can get traction on a word that has been stuck. He may find that an eye blink or an “um” said quickly before trying to say a hard word may short-circuit stuttering temporarily. By the time a child is a teenager, learned reactions influence many of the symptoms. He has learned to anticipate stuttering and may thrash around in a panic when he speaks, trying to escape or avoid it. By adulthood, his fear of stuttering and desire to avoid it can permeate his lifestyle. An adult who stutters often copes with it by limiting his work, friends, and fun to those situations and people that put few demands on speech. Figure 1.1 provides an overview of many of the contributing factors in the evolution of stuttering. In this and the subsequent four chapters, I’ll describe in detail our current understanding of these influences.

Figure 1.1 Factors contributing to the development of stuttering.
Can Stuttering Be Cured?
As implied above, it often cures itself. Some young children who begin to stutter recover without treatment. For others, early intervention may be needed to help the child develop normal fluency and prevent the development of a chronic problem. Once stuttering has become firmly established, however and the child has developed many learned reactions, a concerted treatment effort is needed. Good treatment of mild and moderate stuttering in preschool and early elementary school children may leave them with little trace of stuttering, except perhaps when they are stressed, fatigued, or ill. Most of those who stutter severely for a long time or who are not treated until after puberty make only a partial recovery. They often learn to speak more slowly or stutter more easily and learn to be less bothered by it. Some, however, will not improve, despite our best efforts.
By beginning with a definition of fluency rather than stuttering, I am pointing out how many elements must be maintained in the flow of speech if a speaker is to be considered fluent. It is an impressive balancing act; little wonder that everyone slips and stumbles from time to time when they talk.
Fluency is hard to define. In fact, most researchers have focused on its opposite, disfluency. (I use the term disfluency to apply both to stuttering and to normal hesitations, making it easier to refer to hesitations that could be either normal or abnormal.) One of the early fluency researchers, Freida Goldman-Eisler, showed that normal speech is filled with hesitations (Goldman-Eisler, 1968). Other researchers have acknowledged this and expanded the study of fluent speech by contrasting it with disfluent speech. Dalton and Hardcastle (1977), for example, distinguished fluent from disfluent speech by differences in the variables listed in Table 1.1. Inclusion of intonation and stress in this list may seem unusual. It could be said that speakers who reduce stuttering by using a monotone are not really fluent. We would argue that it is not their fluency but the “naturalness” of their speech that is affected. Nonetheless, both will be of interest to the clinician who works to help clients with all aspects of their communication.
TABLE 1-1 Variables Suggested by Dalton and Hardcastle (1977) as Useful in Distinguishing between Fluent and Disfluent Speech

Starkweather (1980, 1987) suggested that many of the variables that determine fluency reflect temporal aspects of speech production. These include such variables as pauses, rhythm, intonation, stress, and rate that are controlled by when and how fast we move our speech structures. So, our temporal control of the movements of these structures determines our fluency. Starkweather also noted that the rate of information flow, not just sound flow, is an important aspect of fluency. Thus, a speaker who speaks without hesitations but has difficulty conveying information in a timely and orderly fashion might not be considered a fluent speaker.
In his description of fluency, Starkweather (1987) also included the effort with which a speaker speaks. By effort, he means both the mental and physical work a speaker exerts when speaking. This is difficult to measure, but it may turn out that listeners can make such judgments reliably. Moreover, mental and physical effort may reflect important components of what it feels like to be a person who stutters.
In essence, fluency can be thought of simply as the effortless flow of speech. Thus, a speaker who is judged to be “fluent” appears to use little effort when speaking. However, the components of such apparently effortless speech flow are hard to pin down. As researchers analyze fluency more carefully, they may find that the appearance of excess effort may give rise to judgments that a person is stuttering. However, other elements, such as unusual rhythm or slow rate of information flow, may result in judgments that a person is not a fluent speaker, but is not a stutterer either. I will discuss aspects of fluency again when I relate some of the elements of fluency, such as rate and naturalness, to various therapy approaches.
General Description
Stuttering appears at first to be complex and mysterious, but much of it is based on human nature and can be easily understood if you think about your own experiences. In some ways, it is like a problem you might have with your car.
Imagine you had a car that would suddenly stop when you were driving in traffic (Fig. 1.2). Sometimes it would sputter and jerk when you pulled away from a stop sign. Other times, it would drop into neutral, and the engine would race, but the wheels wouldn’t turn. Still other times, the brakes would jam by themselves and wouldn’t release until you stomped repeatedly on the pedal.

Figure 1.2 Stuttering can be like having an old car that often breaks down.
Compare this with what the “core” of stuttering behavior is: Stuttering is characterized by an abnormally high frequency and/or duration of stoppages in the forward flow of speech. These stoppages usually take the form of (a) repetitions of sounds, syllables, or one-syllable words, (b) prolongations of sounds, or (c) “blockages” or “blocks” of airflow or voicing in speech.
Returning to the car analogy: After you’d repeatedly had problems with your car, you would probably develop some coping strategies to get it going again. You might, if it sputtered and jerked, push harder on the gas pedal to try to make it speed up.
Similarly, speakers who are stuttering usually react to their repetitions, prolongations, or blocks by trying to force words out or by using extra sounds, words, or movements in their efforts to become “unstuck” or to avoid getting stuck.
If your car’s problem persisted for several days or longer, you would probably develop some bad feelings about it. The first time it happened, you would be surprised. Then, as it happened more and more, surprise would give way to frustration. If your car frequently quit in the middle of traffic and other drivers nearly hit you and started honking, you would begin to anticipate problems and become afraid they would happen whenever you drove the car.
The child who begins to stutter goes through many of the same feelings–surprise, frustration, embarrassment, and fear. These feelings–in combination with the difficulty the child has in speaking–may cause the stutterer to limit himself in school, social situations, and at work. This might be similar to your responses to a troublesome car. After your car quit on you in traffic many times, you’d probably leave it in the garage and walk, or you’d just stay home.
Another aspect of any description of stuttering involves specifying what it is not. For example, an important distinction must be made between the stuttering behaviors just described and normal hesitations. Children whose speech and language are developing normally often display repetitions, revisions, and pauses–which are not stuttering. Neither are the brief repetitions, revisions, and pauses in the speech of most nonstuttering adults when they are in a hurry or uncertain. Chapter 7 describes the differences between normal disfluency and stuttering in more detail to prepare you for the task of differential diagnosis of stuttering in children.
A distinction should also be made between stuttering and certain other fluency disorders. Disfluency resulting from cerebral damage or disease or psychological trauma differ from stuttering that begins in childhood. In addition, stuttering differs from cluttering, which is another fluency disorder involving rapid, garbled speech that I will talk about in Chapter 15. These disorders may be treated somewhat differently, although some of the techniques that clinicians use with stuttering are also useful with other fluency disorders. These disorders are discussed in Chapter 15.
Core Behaviors
I have adopted the term “core behaviors” from Van Riper (1971, 1982), who used it to describe the basic speech behaviors of stuttering: repetitions, prolongations, and blocks. These behaviors seem involuntary to the person who stutters, as if they are out of her control. They differ from the “secondary behaviors” that a stutterer acquires as learned reactions to the basic core behaviors.
Repetitions are the core behaviors observed most frequently among children who are just beginning to stutter and are simply a sound, syllable, or single-syllable word that is repeated several times. The speaker is apparently “stuck” on that sound and continues repeating it until the following sound can be produced. In children who have not been stuttering for long, single-syllable word repetitions and part-word repetitions are much more common than multisyllable word repetitions. Moreover, children who stutter will frequently repeat a word or syllable more than twice per instance, li-li-li-lilike this (Yairi, 1983; Yairi & Lewis, 1984).
Prolongations of voiced or voiceless sounds also appear in the speech of children beginning to stutter. They usually appear somewhat later than repetitions (Van Riper, 1982), although both Johnson and associates (1959) and Yairi (1997a) reported that prolongations–as well as repetitions–may be present at onset. I use the term prolongation to denote those stutters in which sound or air flow continues but movement of the articulators is stopped. Prolongations as short as half a second may be perceived as abnormal, but in rare cases they may last as long as several minutes (Van Riper, 1982). In contrast to my use of the term, older writers include stutters with no sound or airflow as well as stopped movement of the articulators in their definitions of prolongations (e.g., Van Riper, 1982; Wingate, 1964).
Repetitions and sound prolongations are usually part of the core behaviors of more advanced stutterers, as well as of children just beginning to stutter. Sheehan (1974) found that repetitive stutters occurred in every speech sample of 20 adults who stuttered. Indeed, 66 percent of their stutters were repetitions. Although many of their stutters were also prolongations, as defined above, how many is not clear, because Sheehan’s definition of prolongations seems to differ from mine.
Blocks are typically the last core behavior to appear. However, as with prolongations, some investigators (Johnson and associates, 1959; Yairi, 1997a) have observed blocks in children’s speech at or close to stuttering onset. Blocks occur when a person inappropriately stops the flow of air or voice and often the movement of her articulators as well. Blocks may involve any level of the speech production mechanism–respiratory, laryngeal, or articulatory. There is some evidence and much theorizing that inappropriate muscle activity at the laryngeal level characterizes most blocks (Conture, McCall, & Brewer, 1977; Freeman & Ushijima, 1978; Kenyon, 1942; Schwartz, 1974). Others disagree (Smith, Denny, Shaffer, Kelly, & Hirano, 1996).
As stuttering persists, blocks often grow longer and more tense, and tremors may become evident. These rapid oscillations, most easily observable in the lips or jaw, occur when someone has blocked on a word or sound. The individual closes off the airway, increases air pressure behind the closure, and squeezes her muscles particularly hard (Van Riper, 1982). You can duplicate these tremors by trying to say the word “by” while squeezing your lips together hard and building up air pressure behind the block. Imagine this happening to you unexpectedly when you were trying to talk.
People who stutter differ from one another in how frequently they stutter and how long their individual core behaviors last. Research indicates that a person who stutters does so on average on about 10 percent of the words while reading aloud, although individuals vary greatly (Bloodstein, 1944; Bloodstein & Ratner, 2008). Many people who stutter mildly do so on fewer than 5 percent of the words they speak or read aloud, and a few with severe stuttering stutter on more than 50 percent of the words. The durations of core behaviors vary much less, averaging around one second, and are rarely longer than five seconds (Bloodstein, 1944; Bloodstein & Ratner, 2008).
Secondary Behaviors
People who stutter don’t enjoy stuttering. They react to their repetitions, prolongations, and blocks by trying to end them quickly if they can’t avoid them altogether. Such reactions may begin as a random struggle but soon turn into welllearned patterns. I divide secondary behaviors into two broad classes: escape behaviors and avoidance behaviors. I make this division, rather than follow the traditional approach of dealing with secondary behaviors as “starters” or “postponements,” for example, because my treatment procedures focus on the principles by which secondary behaviors are learned.
The terms “escape” and “avoidance” are borrowed from behavioral learning literature. Briefly, escape behaviors occur when a speaker is stuttering and attempts to terminate the stutter and finish the word. Common examples of escape behaviors are eye blinks, head nods, and interjections of extra sounds, such as “uh,” which are often followed by the termination of a stutter and are therefore reinforced. Avoidance behaviors, on the other hand, are learned when a speaker anticipates stuttering and recalls negative experiences he has had when stuttering. To avoid stuttering and the negative experience that it entails, he often resorts to behaviors he has used previously to escape from moments of stuttering–eye blinks or “uhs,” for example. Or, he may try something different, such as changing the word he was planning to say.
In many cases, especially at first, avoidance behaviors may prevent the stutter from occurring and provide highly rewarding emotional relief from the increasing fear that a stutter will occur. Soon these avoidance behaviors become strong habits that are resistant to change. The many subcategories of avoidances (e.g., postponements, starters, substitutions, and timing devices such as hand movements timed to saying the word) are described in Chapter 7.
When trying to decide if a secondary behavior is an escape or an avoidance, just remember that an escape behavior occurs only after a moment of stuttering has begun, and an avoidance behavior occurs before the moment of stuttering begins.
Feelings and Attitudes
A person’s feelings can be as much a part of the disorder of stuttering as his speech behaviors. Feelings may precipitate stutters, just as stutters may create feelings. In the beginning, a child’s positive feelings of excitement or negative feelings of fear may result in repetitive stutters that he hardly notices. Then, as he stutters more frequently, he may become frustrated or ashamed because he can’t say what he wants to say– even his own name–as smoothly and quickly as others. These feelings make speaking harder as frustration and shame increase effort and tension and impede fluent speech. Feelings that result from stuttering may include not only frustration and shame but also fear of stuttering again, guilt about not being able to help oneself, and hostility toward listeners as well.
Attitudes are feelings that have become a pervasive part of a person’s beliefs. As a person who stutters experiences more and more stuttering, for example, he begins to believe that he is a person who generally has trouble speaking, just as you might believe that your car is a lemon if you continue to have trouble with it. Adolescents and adults who stutter usually have many negative attitudes about themselves that are derived from years of stuttering experiences (Blood, Blood, Tellis, & Gabel, 2001; Gildston, 1967; Rahman, 1956; Wallen, 1960). A person who stutters often projects his attitudes on listeners, believing that they think he is stupid or nervous. Sometimes, however, listeners may contribute directly to the person’s attitudes. Research has shown that most people, even classroom teachers and speech-language pathologists, stereotype people who stutter as tense, insecure, and fearful (e.g., Turnbaugh, Guitar, & Hoffman, 1979; Woods & Williams, 1976). Such listener stereotypes can affect the way individuals who stutter see themselves, and changing a client’s negative attitudes about himself can be a major focus of treatment.
The three components of stuttering–core behaviors, secondary behaviors, and feelings and attitudes–are depicted in Figure 1.3. The core behavior is the individual’s block on the “N” in “New York.” The secondary behaviors consist of postponement devices such as “uh,” “well,” and “you know” and substitution of “the Big Apple” for “New York.” Feelings and attitudes are depicted as the individual’s thoughts that he won’t succeed in saying the word fluently and the individual’s belief that listeners will think he is dumb because he stutters.

Figure 1.3 Components of stuttering: core behaviors, secondary behaviors, and feelings and attitudes.
Functioning, Disability, and Health
Some time ago, the World Health Organization (WHO) adopted the International Classification of Impairment, Disabilities, and Handicaps (WHO, 1980) to describe the consequences of various diseases and disorders. A number of authors have applied this framework to stuttering (Curlee, 1993; McClean, 1990; Prins, 1991, 1999; Yaruss, 1998, 1999). A decade ago, WHO changed their taxonomy to the International Classification of Functioning, Disability, and Health (ICF) (WHO, 2001). In the following paragraphs, I will suggest ways in which this system may be applied to stuttering.
The taxonomy begins with “Functioning and Disability,” wherein body structures and body functions are considered. Structures that are dysfunctional in stuttering, as brain imaging studies have shown, are cortical and subcortical structures, such as white matter tracts that may be critical for coordinating planning, execution, and sensory feedback for speech. Functions that differ in stuttering are the interruptions of speech flow that characterize the disorder. The ICF system becomes more useful when “Activity and Participation” are considered. Individuals who stutter may be affected to a greater or lesser extent in two of the ICF areas, “Speaking” and “Conversation.” These are domains in which stuttering is noticeable. A third area, “Interpersonal Interactions,” may also be affected if speaking and conversation are restricted by the stuttering to the extent that the person who stutters refrains from fully engaging with others.
A new and important section of the latest ICF system is titled “Contextual Factors.” One component of this section is “The Environment.” This is particularly relevant to individuals who stutter because people in the environment may range from unsupportive (e.g., a home with great stress or classmates who tease a child) to highly supportive (e.g., a family that is accepting of the child and encouraging of her participation). Also under “Contextual Factors” is the category of “Personal Factors.” These are the attributes of a person who stutters–her character and personality.
Consider the influence of environmental and personal factors on two individuals who stutter. The first is the successful former CEO of General Electric, Jack Welch, who authored Jack: Straight from the Gut. His assertive temperament and early acceptance of his stuttering by his family were no doubt important in helping him succeed in the high-pressure world of corporate boardrooms. From an early age, Welch refused to let stuttering stand in the way of his goals (Welch & Byrne, 2001). In contrast, actor James Earl Jones initially reacted to his stuttering in a vastly different way. When he was 6 years old, he was so traumatized by his stuttering, he pretended that he was mute so that he wouldn’t have to speak. Only later, with the support of someone in his environment–a high school English teacher–did he begin to learn that he could overcome his stuttering by facing difficult situations and practicing reading aloud in front of an audience (Jones & Niven, 1993).
Another two examples come to mind–men who had stuttered severely since childhood but obtained excellent college educations, were highly successful in business, and used their wealth to help others. One is Malcolm Fraser, who was a cofounder of the National Auto Parts Association and created the Stuttering Foundation of America. The other is Walter Annenberg, who established a media empire and later the Annenberg Foundation, a large philanthropic organization.
In all four cases, their functioning may have been impaired, but environmental and personal factors enabled them to overcome potential limitations in the domains of speaking and interpersonal interactions. You can see in this classification system why clinicians play a vital role in the lives of children and adults who stutter. They can influence environmental factors by helping families, teachers, and entire schools become supportive of the individuals who stutter and facilitative of increased fluency. And they can build the personal attributes of each client through counseling, insightful listening, educating, and caring.
Before I delve deeper into the basic facts about stuttering, I’d like to touch briefly on the personal side of the problem. Some of you may never have had a friend who stutters or may never have worked with a stutterer in treatment, so I will present several examples of what stuttering can be like. Even if you are familiar with stuttering, these brief sketches, which portray four individuals who differ in age and in their accommodations to stuttering, may expand your sense of what stuttering is like for the person who experiences it. These case studies begin on page 12. You may also visit thePoint to watch video clips of these different levels of stuttering.
This section relates some of the best-known “facts” about stuttering. These are replicated research findings that pertain to the occurrence and variability of stuttering in the population and in individuals. As we discuss these findings, we will note what they suggest about the nature of stuttering. Thus, as you read the rest of this chapter, you will become increasingly aware of my perspective on the nature and treatment of stuttering.
Much has been made of the “heterogeneity” of stuttering; a number of authors have suggested that stuttering is not one disorder, but many. Researchers have proposed various divisions of the disorder, such as Van Riper’s (1982) four “tracks” of stuttering development and St. Onge’s (1963) triad of speech-phobic, psychogenic, and organic stutterers. My approach is to focus on the majority of people who stutter –those whose stuttering begins during childhood without an apparent link to psychological or organic trauma. This most common type of stuttering has been called “developmental stuttering,” because symptoms usually emerge gradually as a child develops, especially during the period of intense speech and language acquisition. I simply call it “stuttering.” In denoting similar fluency problems that are associated with psychological problems, brain damage, cognitive impairment, and cluttering, I refer to their assumed etiology, such as “disfluencies associated with brain damage.”
Note, however, that even within the group of individuals whose stuttering begins in early childhood during rapid speech and language development, there is a great deal of variability in the behaviors we call stuttering and in how these behaviors change (or don’t) as the child progresses toward persistence or recovery.

Case Example

Katherine developed speech and language normally, speaking her first word at about 1 year and beginning to combine words at 15 months with complete fluency. When she was 3, after a particularly hectic Christmas holiday, she began to stutter. Her first disfluencies were easy part- and whole-word repetitions, but she soon began to tense her articulators, momentarily blocking the flow of speech until the word “popped out.” Sometimes, when she was completely stuck for several seconds, she responded by hitting her parents or crying out. She also showed much less interest in talking and using new words and phrases.
Her parents soon brought her to a speech and language clinic for an evaluation. Katherine was found to be stuttering on 21 percent of the syllables she spoke–a very high percentage for any child. Her overall severity was assessed with the Stuttering Severity Instrument, which rated it as severe. Her receptive language was found to be far above average for her age. Her expressive language was found to be typical for her age, but it was likely that she was inhibited in expressing herself because of her stuttering. Her phonological development was found to be appropriate for her age.
For a description of Katherine’s treatment, see Chapter 12.
In Segment 1, Katherine responds to a question from her mother by saying “O-OO-O…an…an…an…OK now…I think he is still hungry…[unintelligible]…a leaf.” This may be an example of a child getting stuck on an attempt to say a word (“OK”) and then, finding herself in a block, changing the word she tries to say (from “OK” to “and”). Then she is able to say the original word and go on.
In Segment 2, still with her mother, she is stuck on the first sound of the word “OK,” then she seems to be able to move on to the “kay” but gets stuck there too, and so she doubles back to the “o” and finally finishes the word by pushing out of the stutter on “o” and then having a slight stutter on the first sound of “kay.” Whew! You can see how much work this 3-year-old must do just to get a few words out. No wonder she doesn’t seem as expressive as she did before her stuttering started.
In Segment 3, Katherine is playing with the clinician and says “uh…nnnnn… ne…ne…nnnn…na…now, what is this?” The “uh” might be Katherine’s way of getting ready for the stutter she anticipates on “now.”
In Segment 4, Katherine is playing with both of her parents and says “Lo-look what I made. Oh…uh…bbbbb…buh…bbbbb…buh…” The first stutter is a partword repetition (“lo-look”) that is so mild it might be considered a normal disfluency in another context. But the second stutter (on a word beginning with “b”) is quite a long stutter in which Katherine struggles heroically to produce the word but is interrupted by the door opening before she can finish. How would you describe this last stutter? Repetition? Prolongation? Block? What emotions do you think she’s feeling? Why do you think so?

David was the second of three children in a family with no history of speech or language disorders. His speech was developing normally until age 4, when he began to show excessive part-word and whole-word repetitions. After several months, when David’s stuttering had not decreased, his mother took him to his pediatrician who assured her it would resolve on its own.
When David was almost 6, his stuttering was growing steadily more severe, and he was avoiding talking in many situations. His mother then decided to consult a speech-language pathologist at a university clinic, who evaluated David. In the evaluation, David was stuttering on 8 percent of his syllables spoken; many stutters were tightly squeezed blocks with evident struggle behavior.
David’s subsequent treatment and his current status as a 20-something-year-old are described in Chapter 13.
In this video, you can see how stuttering may be more complicated as children grow older and become more self-conscious. In the video for the first child, Ashley’s moments of stuttering seemed to pop up out of the blue and surprise her. In the video for the second child, Katherine’s stuttering was a little more predictable to her, but it was mostly confined to those few words on which she got blocked. Now we will see that David’s stuttering affects more of his entire speech pattern and is characterized by much avoidance and struggle.
In Segment 1, showing David talking with his mother, David’s style of speaking is hesitant, with many stops, “ums,” false starts, and changes in direction. He says something like “and then…it like that…and then…then put… um the same a-mount of…of…of…of …um… [then some unintelligible words, after which he seems to give up on the sentence and begins counting].” As you can imagine, it’s hard to assess what percentage of syllables are stuttered when can imagine, it’s hard to assess what percentage of syllables are stuttered when David avoids saying the words on which he expects to stutter.
Segment 2 shows a block on the word “whoever,” preceded by several words that seem to postpone David’s attempt on this word. He says “and then …uh … whoever gets um …um the four, four… an, an … um … um these [unintelligible word].” Do you think he expects to stutter on “whoever?” What are the cues that tell you so? What escape behaviors does David use as he struggles with this block?
In Segment 3, David has another block with a few avoidance behaviors before and escape behaviors during the stutter. He says “He … [unintelligible word] … he goes home automatically because … um … because he-he has done the shortcut an-an he goes all the way home.” Can you describe the avoidance and escape behaviors David shows in this clip?
An Adult: Advanced Stuttering

Sergio is a 44-year-old musician who has stuttered since he was 3 years old. Eight of his maternal aunts and uncles stuttered, suggesting a genetic origin to his problem. His stuttering began as multiple repetitions of one-syllable words and parts of words. Much of his speech was fluent, but whenever he was excited or hurried, Sergio’s stuttering flared up, sending his parents into a state of alarm and concern for his future. At first, his father’s solution for Sergio’s stuttering was hitting him on the head with his knuckles when he blocked. When this failed and Sergio developed physically tense prolongations and blocks that occurred regularly in his speech, his parents took him to various therapists, including a hypnotist and a psychotherapist who prescribed tranquilizers. None of these seemed to have more than a temporary effect, and Sergio’s stuttering grew steadily worse. During his elementary and junior high school years, he was frequently ridiculed for his stuttering, even by teachers, and Sergio found himself This changed, however, soon after “Beatlemania” swept through America. Sergio bought a guitar and taught himself to sing “I Want to Hold Your Hand” and other Beatles songs. As a result, his popularity with schoolmates shot up, even though his stuttering continued to worsen. He had so much difficulty speaking in class, and his teachers were so unsympathetic, he finally dropped out of school and pursued a vagabond lifestyle as a singer and songwriter.
As he traveled, working various jobs by day and singing at night, Sergio continued to stutter severely, with one happy exception. When he was performing with his band, not only did he sing fluently, but he also spoke to the audience easily, announcing each number and making casual, funny comments between songs. As a result of his constant battle with stuttering, Sergio developed a wide variety of avoidances. He dodged making phone calls, and whenever he received calls, he used elaborate facial grimaces and starter sounds to fight his way through stutters.
The telephone is a difficult situation for most people who stutter, and Sergio is no exception. Segment 1 shows Sergio talking about his experiences on the phone when he had long silent blocks in the past, and Segment 2 is a phone call Sergio made more recently. On both clips you’ll see a mix of avoidance and escape behaviors that are now well entrenched in Sergio’s speech pattern after years of stuttering. There are also some straightforward stutters without escape and avoidance behaviors that are witness to Sergio’s attempts to stutter in a simpler way. See if you can identify the types of stutters that Sergio has on both segments, as well as his particular escape and avoidance behaviors, more of which are seen in Segment 2.
Imagine yourself in your doctor’s office with an annoying cold that just won’t go away –runny nose, sore throat, and cough. She asks you to describe when the first signs of your illness appeared and what they were like at onset. It is quite likely you won’t remember exactly when your symptoms first occurred and exactly what they were like, especially if they came and went over the course of a week or two before they became persistent. This is the problem with determining the onset of stuttering. Parents are asked to recall exactly when the child’s stuttering started and what it was like when they first noticed it; thus some of our information on onset–especially from older studies–may be inaccurate. The description of stuttering onset given here is relatively brief. More details are given in Chapter 7 when I describe the differences between normal disfluency and the beginning stages of stuttering.
Let’s first consider the question of how old children are, on average, when they begin to stutter. In the earliest studies (e.g., Milisen & Johnson, 1936), researchers asked parents a year or more after the onset of stuttering had occurred, to recall the age of stuttering onset in their children. The average age of onset, taken from nine pre1990 studies summarized in Bloodstein and Ratner (2008), is roughly 4 years. After 1990 studies summarized in Bloodstein and Ratner (2008), is roughly 4 years. After 1990, led by Ehud Yairi and his colleagues at the University of Illinois, researchers were careful to interview parents of children who began to stutter within 12 months of the parent interview. Bloodstein and Ratner (2008) list six studies conducted after 1990 in which parents were interviewed closer to onset than earlier studies. The average age of onset of these newer studies is about 2.8 years. Thus, either the newer studies are getting a more accurate picture of the age of onset, the age of onset is getting younger, or both.1 The current consensus is that the onset of stuttering typically occurs just before age 3, and most onsets occur between ages 2 and 3.5 years (Yairi & Ambrose, 2005). Some older children–up to about age 12–may begin to stutter, but these are much rarer cases. Stuttering onset in adolescents and adults is likely to be a different form of disfluency–psychogenic or neurogenic–which I will discuss in Chapter 15.
Next, let’s look at the first signs of stuttering, as reported by parents. Most early reports of stuttering onset (e.g., Bluemel, 1932) indicated that simple, relaxed repetitions of syllables and words were the typical first signs of stuttering. However, some early studies (e.g., Taylor, 1937) and the carefully conducted interviews by Yairi (1983) found that, in many cases, parents described prolongations and blocks, along with signs of struggle, as the first stutters shown by their child. Summarizing their own and others’ research, Yairi & Ambrose (2005) suggest that even when only repetitions are the first signs of stuttering, the percentage of spoken syllables that are repeated and the number of iterations in each repetition are markedly higher in children who stutter than in their normal peers.
Lastly, there is the question of whether onset is sudden, intermediate, or gradual. In other words, do parents conclude their child is stuttering because of marked disfluencies seen in the course of a day or two, does it take a week or two for them to make that determination, or does it take longer, many weeks? Remember our example of getting a cold and trying to remember the onset? No doubt some colds come on suddenly, with sore throat and running nose appearing overnight and getting worse quickly. Other colds tiptoe into your life, with a sore throat that comes and goes and later turns into a runny nose and cough.
In contrast to the earliest reports on stuttering always having a gradual onset, Yairi and Ambrose (2005) found in their sample of 163 children many cases (41 percent) in which onset was reported to be sudden, with another group (32 percent) reported as intermediate, and a third group (27 percent) as gradual. These figures may be influenced by how attentive to their child’s speech these parents were; no doubt, parents who had relatives who stuttered would have been more likely to recognize stuttering sooner.
The term “prevalence” is used to indicate how widespread a disorder is. Information about the prevalence of stuttering tells us how many people currently stutter. Accurate, up-to-date information on the prevalence of stuttering is difficult to obtain. The research literature contains studies having many methodological differences, which of stuttering probably varies considerably with age, and not all studies measure stuttering in the same age groups. Moreover, definitions of stuttering may vary from study to study. Some studies may include relatively normally disfluent individuals in their count; others may exclude them.
Beitchman, Nair, Clegg, and Patel (1986) assessed the prevalence of speech and language disorders in kindergarten children, using a representative sample. They retested children who failed the initial screening as well as a random sample of children who passed. The prevalence of stuttering in this sample of kindergarten children was 2.4 percent. Although this is only one study’s finding, the care with which the data were collected increases its credibility.
Bloodstein and Ratner (2008) reviewed and summarized the results of 44 studies of school-age children in the United States, Europe, Africa, Australia, and the West Indies. These studies showed that the prevalence of stuttering throughout the school years is about 1 percent. Andrews, Hoddinott, Craig, Howie, Feyer, and Neilson (1983) came to the same conclusion–about 1 percent of the schoolchildren worldwide are likely to be stutterers at any given time. If the 2.4 percent prevalence among kindergartners noted above is valid, a considerable number of recoveries must take place between kindergarten and the upper grades.
There appear to be no reliable prevalence data for stuttering in adults. However, both Andrews and colleagues (1983) and Bloodstein and Ratner (2008) suggest that the prevalence of stuttering is lower after puberty. If so, the prevalence for adults would be less than 1 percent.
The incidence of stuttering is an index of how many people have stuttered at some time in their lives. Like the data on prevalence, incidence figures are not clear-cut because different researchers have used different definitions of stuttering and methods for obtaining their data. Some researchers only report stuttering that lasted six months or more, not wanting to include shorter episodes of disfluency. Others report any speech behaviors that informants or parents considered to be stuttering. Estimates of incidence, when reports of informants and parents are considered, are as high as 15 percent, a figure that includes those children who stuttered for only a brief period (Bloodstein & Ratner, 2008). When only the cases of stuttering that lasted longer than six months are included, incidence appears to be about 5 percent (Andrews et al., 1983). We think the latter estimate may more accurately reflect the chronic disorder we call stuttering, but the former illustrates how close perceptions of normal disfluency and early stuttering may be.
A report by Mansson (2000) of all children born on the Danish island of Bornholm supports the suggested incidence of 5 percent. This population is homogeneous and stable, making a careful, longitudinal study quite possible. Of the 1,042 children born on the island in 1990 and 1991, 98 percent were screened for speech and language problems at age 3. The children were followed for nine years, and it was then found that 5.19 percent went through some period of stuttering.
Incidence figures tell us something else about the nature of stuttering. The Incidence figures tell us something else about the nature of stuttering. The difference between incidence (5 percent) and prevalence (1 percent in school-age children and less in adults) suggests that most people who stutter at some time in their lives recover from it, and we know that prevalence declines after puberty. Thus, unless treatment alone is responsible for such remissions, some aspect of growth or maturation allows many individuals to recover from stuttering.
Recovery without Treatment
Recovery from stuttering without treatment, also referred to as “spontaneous” or “natural” recovery, has long been a puzzling issue. Putting aside the important question about why children recover without treatment, there is debate about what percentage of children who start to stutter recover in this way.
Reviews of early research report findings that range from 20 percent to 80 percent natural recovery (Bloodstein & Ratner, 2008; Andrews et al., 1983). This wide range may be from different methodologies used by different studies. Some asked large numbers of adults if they ever stuttered when they were children. This method, which is called “retrospective” may be affected by faulty memories, poor definitions of stuttering, and the inclusion of individuals who may have stuttered for only brief periods.
Recent research has proceeded more carefully by first identifying a group of children close to the onset of their stuttering and then following them for several years without offering treatment and assessing how many recover and how many persist. Those who persist are then referred for therapy. Several studies using this methodology have been published. Yairi and Ambrose (1999) followed a group of 84 children for a minimum of four years after the onset of their stuttering and determined that over this span of time, 74 percent had recovered without treatment. Kloth, Kraaimaat, Janssen, and Brutten (1999) followed 23 children for six years and discovered that 70 percent had recovered. Mansson (2000) identified 51 children between the ages of 3 and 5 who started to stutter and found that 71 percent recovered within two years. When the follow-up continued for another few years until the children were 8 or 9 years old, recoveries were up to 85 percent. However, this latter figure may be affected by speech therapy given to some after the two-year follow-up.
Several studies have compared children who recover and those who persist to determine what might characterize children who recover. Research at the University of Illinois (Yairi & Ambrose, 2005) over the past 20 years indicates that there are several factors that are useful for indicating the likelihood that a child’s stuttering will persist rather than disappear naturally. You may wish to follow up on this brief overview by reading the reference, given above, that describes these findings in detail. The following factors appear to be among the most important predictors:
Family history: When a child’s family includes stutterers whose stuttering persisted, there is increased risk of persistence.
Gender: Boys have a greater risk of persistence. However, girls typically recover more quickly; therefore, a girl who has been stuttering for a year is at more risk than a boy who has been stuttering for a year.
Age at onset: Children who begin to stutter “later” have a greater risk of persistence. Onsets occur most frequently between ages 2 and 3.5 years, so children with onset after 3.5 years are more at risk.
Trend of stuttering frequency and severity: Children whose stuttering (defined as part-word repetitions and single-syllable word repetitions, prolongations, and blocks) frequency and severity is not decreasing over a period of a year after onset are at more risk of persistence.
Duration since onset: The longer the child continues to stutter beyond a year after onset, the greater the risk of persistence, especially for girls.
Duration of stuttering moments: Continued presence of more than one repetition unit, especially more than three (li-li-li-li-like this) is a sign of increased risk. Also, continued rapid repetitions are a sign of increased risk. Children who recover tend to have fewer repetition units (li-like this) and slower repetitions (li…….like this).
Continued presence of sound prolongations and blocks: The percentage of prolongations and blocks at onset doesn’t predict persistence, but if prolongations and blocks do not decrease as stuttering goes on, the child is more likely to persist.
Phonological skills: Children whose phonological skills are below the norms have a greater risk for continued stuttering.
Two other studies examined factors associated with recovery. A longitudinal study by Brosch, Haege, Kalehne, and Johannsen (1999) followed a group of 79 stuttering children for several years. The group that persisted in stuttering had a significantly larger proportion of left-handed children. Because this is a preliminary report from an ongoing study, caution should be exercised in considering this factor as critical to recovery. Nonetheless, the factor of laterality may be one of the additional genetic factors that influence recovery; replication of this work is critical.
In the study by Kloth et al. (1999) described earlier, results indicated that children who recovered had a more mature (less variable) speech motor system, a slower speaking rate, and a mother whose interaction style was nondirective and whose language was less complex. Rommel, Hage, Kalehne, and Johannsen (2000) followed 71 children identified as stuttering soon after onset and followed them for three years. The mothers of those who recovered naturally compared to the mothers of those who persisted had less complex syntax and a smaller number of different words used when talking to their children.
In summary, early studies of recovery reported wide variations in results. Their findings depend on many factors–among them, the accuracy with which stuttering is differentiated from normal disfluency, whether the study is retrospective or longitudinal, and the size of the group studied. The most careful studies are longitudinal assessments of children who are identified soon after the onset of stuttering and are followed for several years. In these studies, about 75 percent of the children who begin to stutter recover without formal treatment. Many factors have been suggested as associated with recovery. These include having relatives who recovered from stuttering or no relatives who stuttered; having an early onset; showing a decrease in frequency and severity of stuttering in the year after onset; having a slower speech rate; having a more stable speech-motor system; having a mother who has a nondirective interaction style and uses less complex language when speaking to the child; being right-handed; having less severe stuttering (although some studies dispute this); having good phonological, language, and nonverbal skills; and being female. We now consider the latter variable, the sex factor, in more detail.
Sex Ratio
Studies of the sex ratio in stuttering were first published in the 1890s and have been published every decade since. With this steady stream of information, we ought to have reliable data on this phenomenon. In fact, we do. The results from studies of people who stutter at many ages and in many cultures put the ratio at about three male stutterers to every one female stutterer. There is strong evidence, however, that the ratio may increase as children get older. For example, Yairi (1983) reported that of 22 children who were 2 and 3 years of age and whose parents believed they were stuttering, 11 were boys, and 11 were girls. In a larger study of 87 children between 20 and 69 months, Yairi and Ambrose (1992b) found a male:female ratio of 2.1:1 overall, although the 20 youngest subjects, those under 27 months, showed a 1.2:1 ratio.
Bloodstein and Ratner’s (2008) review indicated that the male-to-female sex ratio is about 3:1 in the first grade and 5:1 in the fifth grade, confirming the hypothesis that the sex ratio increases as children get older. Evidence of the increasing male-to-female ratio was provided by two recent studies. Kloth and colleagues (1999) found a male:female ratio of 1.1:1 ratio near onset, which rose to 2.5:1 six years later. Mansson (2000) found a male:female ratio of 1.65:1 at the initial screening (age 3), which rose to a ratio of 2.8:1 two years later. The nearly even sex ratio among very young children who stutter and the gradually increasing proportion of boys who stutter may be a consequence of several factors. West (1931) presented data indicating that the change in sex ratio was the result of an increasing proportion of boys beginning to stutter in the late preschool and early school-age years. However, recent data indicate that girls begin to stutter a little earlier than boys (Yairi, 1983; Yairi & Ambrose, 1992b) and recover earlier and more frequently (Andrews et al., 1983; Yairi & Ambrose, 1992b; Yairi & Ambrose, 1999; Yairi, Ambrose, & Cox, 1996).
Females who stutter and don’t recover by adulthood may be an interesting subpopulation to study. They may have inherited a stronger predisposition to stutter, have been subjected to strong environmental pressures on their speech, or both (Andrews et al., 1983). Alternately, they may lack the “recovery factor” that most young female stutterers appear to have, or they may have inherited additional factors that interact with stuttering to inhibit recovery.
Variability and Predictability of Stuttering
Another important piece of background information about stuttering is how it varies yet is surprisingly predictable in its occurrence despite the fact that it seems so inconsistent and so idiosyncratic. This predictability is an important clue to its nature.
As we trace the research on stuttering’s variability, we will see how this information reflects changing theoretical perspectives on the disorder.
Before the 1930s, stuttering had been commonly regarded as a medical disorder. Lee Edward Travis, the first person trained as a Ph.D. to work with speech and hearing disorders, set up a laboratory at the University of Iowa in 1924 to study stuttering from a neurophysiological perspective. He hypothesized that stuttering was the result of an anomalous or inefficient organization of the brain’s two cerebral hemispheres. To Travis and his fellow researchers, the variability of stuttering behaviors was seen as part of an organic disorder, and an unimportant part at that. Far more relevant to their research were stutterers’ brain waves, heart rates, and breathing patterns. But in the 1930s, psychologists at Iowa and elsewhere began taking a keen interest in behavioral approaches to the study of human disorders, which spilled over into research on stuttering. Scientists who had been trying to understand the neurophysiology of stuttering gradually began trying to examine the social, psychological, and linguistic factors that govern its occurrence and variability (Bloodstein & Ratner, 2008).
Anticipation, Consistency, and Adaptation
Before describing these interesting findings, I’ll briefly explain these terms that are best understood in the context of someone who stutters reading a passage several times. “Anticipation” refers to an individual’s ability to predict what words or sounds he or she will stutter on (Johnson & Solomon, 1937; Knott, Johnson, & Webster, 1937; Milisen, 1938; Van Riper, 1936). “Consistency” is the tendency for people to stutter on the same words when they read a passage more than once (Johnson & Inness, 1939; Johnson & Knott, 1937). “Adaptation” is the finding that when speakers read a passage several times, they gradually stutter less and less over the course of five or six readings (Johnson & Knott, 1937; Van Riper & Hull, 1955).
These studies were usually carried out by giving the stutterer a passage and asking him to read it aloud. If the experimenter were studying consistency, for example, he would have his own copy of the passage on which he would mark every word on which the speaker stuttered. Then he would ask the speaker to read it again, and the experimenter would again mark the words stuttered in the second reading. From this he could calculate the percentage of words stuttered in two (or more) readings. These findings, called anticipation, consistency, and adaptation, respectively, changed some assumptions about the disorder. Stuttering, it seemed, was not simply a neurophysiological disorder. It showed characteristics of learned behavior, as well.
These studies not only changed existing views of stuttering, they also opened the door to new treatment possibilities. If much of stuttering is learned, it may be unlearned. The challenge was to determine how much is learned and how to help people who stutter develop new responses. Many of the treatment approaches we discuss later in the book use principles of learning to help clients acquire new, fluent responses and reduce the tension and avoidance of their old stuttering responses.
Language Factors
One of the Iowa researchers, Spencer Brown, pushed investigations of the predictability of stuttering into the realm of language. In seven studies completed over a stretch of 10 years, Brown found correlations between stuttering and seven grammatical factors during reading aloud. These findings were reported in a remarkable series of papers Brown published from 1935 to 1945 (Brown, 1937, 1938a, 1938b, 1938c, 1943, 1945; Brown & Moren, 1942; Johnson & Brown, 1935). Brown showed that most adults who stutter do so more frequently
• On consonants
• On sounds in word-initial position
• In contextual speech (vs. isolated words)
• On nouns, verbs, adjectives, and adverbs (vs. articles, prepositions, pronouns, andconjunctions)
• On longer words
• On words at the beginnings of sentences
• On stressed syllables
These findings strongly suggest that stuttering is highly influenced by these linguistic factors.
Later investigators applied Brown’s hypotheses to the speech of children who stutter. An advantage in studying language factors in children’s stuttering is that the loci (places where it occurs in speech) and frequency of stuttering might be less influenced by responses learned from years of stuttering and more by innate language processing difficulties. Indeed, researchers discovered that although stuttering in elementary school children follows the same linguistic patterns as adult stuttering, the loci and frequency of stuttering in preschool children are different. Stuttering in these very young children occurs most frequently on pronouns and conjunctions, not on nouns, verbs, adjectives, and adverbs. It occurs not as repetitions, prolongations, or blocks of sounds in word-initial positions, but as repetitions of parts of words and single-syllable words in sentence-initial positions (Bloodstein & Gantwerk, 1967; Bloodstein & Ratner, 2008). This led researchers to hypothesize that in its incipient stage, stuttering is located at the beginning of syntactic units (sentences, clauses, and phrases), as if the task of linguistic planning and preparation was a key ingredient in the recipe for disfluency (Bernstein Ratner, 1997; Bloodstein, 2001, 2002; Bloodstein & Ratner, 2008).
Conture (2001) and others (e.g., Byrd, Wolk, & Davis, 2007) have focused particular attention on the phoneme or sound selection component of linguistic planning in individuals who stutter. Findings that recovery from stuttering may be associated with good phonological skills, a slower speech production rate, and a stable speech-motor system suggest that some individuals who stutter may overcome a linguistic planning delay by relying on strengths in related language areas or by slowing their rates of speech production to compensate for such deficits. We will revisit these modes of compensation when we discuss treatment approaches.
In summary, there are strong links between language and stuttering. As mentioned
in the section describing onset, stuttering usually first appears when children are going through the most intense period of language acquisition (Bloodstein & Ratner, 2008). It is also clear that deficits in language (including phonology) often accompany stuttering and may predict its persistence (Yairi & Ambrose, 1999, 2005). Future studies may also confirm what several researchers have suggested, that even stutterers who show no clinically significant language disorders may have subtle language or phonological deficits that contribute to their stuttering (e.g., Byrd, Wolk, & Davis, 2007).
Fluency-Inducing Conditions
One of the researchers at the University of Iowa, Oliver Bloodstein, wrote his Ph.D. dissertation on “Conditions under Which Stuttering Is Reduced or Absent” (Bloodstein, 1948, 1950). In studying the speech of stutterers in 115 conditions, Bloodstein found that stuttering is markedly decreased in many of these conditions. Some of these conditions are speaking when alone, when relaxed, in unison with another speaker, to an animal or an infant, in time to a rhythmic stimulus or when singing, in a different dialect, while simultaneously writing, and when swearing. In later studies, reviewed in Andrews, Howie, Dosza, and Guitar (1982), additional conditions were found to reduce stuttering. These conditions included speaking in a slow prolonged manner, speaking under loud masking noise, speaking while listening to delayed auditory feedback, shadowing another speaker (repeating what they say immediately afterward), and speaking when reinforced for fluent speech.
Various explanations have been proposed to account for the impact of these conditions. Most are compatible with the idea that stuttering has a substantial learned component and is affected by such external stimuli as communicative pressure. Recent brain imaging studies–reviewed in Chapters 2 and 3–indicate that left-hemisphere speech and language structures and functions are impaired in people who stutter. This may make it difficult for a speaker to orchestrate rapid and coordinated production of phonological and lexical items, syntax, intonation, and other subcomponents of spoken language. Thus, many conditions may induce fluency by providing timing cues, reducing rate, lowering stress on vulnerable pathways, or marshaling attentional resources to overcome the limitations of an impaired neurological system.
An Integration
Modern research on stuttering has taken a long and complex journey from Travis’s laboratory in Iowa in 1924. Yet, in many ways, those early findings are not entirely irrelevant. Travis’s theory of stuttering as a problem of coordinating the two sides of the brain for speech has reemerged as a view of stuttering as a problem of coordinating multiple brain networks for speech with those for language, cognition, and emotion. This juggling act breaks down in all speakers when the resources needed to process language, cognition, or emotion momentarily drain available central nervous system capacities, leaving too little capacity for the intricacies of rapid, smooth speech production. The result is normal disfluency. Those individuals who stutter appear to have trouble allocating resources to speech production under conditions of high demand. They have inherited or acquired a more vulnerable speech production system –one that is less able to deal with the norm of rapid, smooth speech under a wide variety of conditions, perhaps because the neural pathways used to produce speech are not as efficient as they need to be.
This vulnerable speech production system may heal itself and become more resistant to disruption in children who recover naturally from stuttering. Because of that great neural plasticity characteristic of the very young, some of these children’s brains spontaneously develop new, more efficient pathways for speech and become entirely fluent. Others may recover because they learn to compensate by speaking more slowly or finding other ways to marshal their resources to resist disruption. The children who don’t recover appear to fall into a cycle of reacting to their disfluencies by tensing muscles, struggling to escape, and even avoiding difficult words and situations. These highly learned reactions–which are influenced by an individual’s personality and the people around him–become part of an individual’s stuttering patterns and influence the way he thinks and feels about speaking.
This updated view is essentially the model of stuttering presented in this book. To state it more formally, stuttering is an inherited or congenital disorder that first appears when a child is learning the complex and rapid coordinations of speech and language production. Children who do not recover but persist in stuttering are those who may have more extensive deficits in the neural networks of the brain. As stuttering persists, they learn maladaptive responses to their disfluencies. This learning is influenced by their biological temperament, developing social and cognitive awareness, and the response of the environment to their speech.
The next few chapters expand on this theme and prepare you to use this information in diagnosis and treatment.

• Stuttering appears in all cultures and has been a problem for humankind for at least40 centuries.
• It is characterized by a high frequency or severity of disruptions that impede theforward flow of speech.
• It begins in childhood and usually becomes more severe as the child grows to adulthood unless he recovers with or without formal treatment.
• Core behaviors of stuttering are repetitions, prolongations, and blocks. Secondarybehaviors are the result of attempts to escape or avoid core behaviors and include physical concomitants of stuttering, such as eye blinks or verbal concomitants, such as word substitutions.
• Feelings and attitudes can also be important components of stuttering that reflect thestutterer’s emotional reactions to the experience of being unable to speak fluently and to listener responses to her stuttering. Feelings are immediate emotional reactions and include fear, shame, and embarrassment. Attitudes crystallize more slowly from repeated negative experiences associated with stuttering. An example is a stutterer’s belief that listeners think she is stupid when they hear her stuttering.
• Stuttering begins between 18 months of age and puberty, but most often betweenages 2 and 5 years, with a peak just before age 3. Its first appearance may be either a gradual increase in easy repetitions of words and sounds or a sudden onset of multiple repetitions, sometimes with prolongations or blocks as well.
• Prevalence of stuttering is about 1 percent. Incidence is about 5 percent. Recoveryrate without professional treatment is about 75 percent of children who ever stuttered. The male-to-female ratio in schoolchildren and adults is about 3:1 but may be lower, close to 1:1, in very young children who start to stutter. More girls recover during early childhood, increasing the proportion of males with the disorder after the preschool years.
• Many persons who stutter are able to predict the words in a reading passage theywill stutter on before reading it aloud (anticipation), and most tend to stutter on many of the same words each time in repeated readings of a passage (consistency). Stuttering frequency decreases for most stutterers when they read a passage many times (adaptation).
• Stuttering occurs more frequently in certain grammatical contexts. The nature ofthese grammatical contexts differs somewhat for adults and children.
• A variety of conditions reduces the frequency of stuttering. Their effects may beattributable to changes in speech pattern, reductions in communicative pressure, or both. Research on these fluency-inducing conditions suggests that stuttering may be decreased by conditions that reduce the demands on speech motor control and language formulation functions.

1. What might make some children’s core behaviors change from repetitions toprolongations to blocks?
2. What are the differences between core and secondary behaviors in stuttering?
3. When stuttering is defined, from what other kinds of hesitation must it be distinguished?
4. What are some feelings and attitudes people who stutter might have, and what istheir origin? Do nonstutterers ever have these feelings?
5. What is the age range for the onset of stuttering (the youngest and oldest ages atwhich onset is commonly reported)? Why might it occur at that time?
6. What is the difference between “incidence” and “prevalence?”
7. What problems do researchers encounter when they try to determine how manystutterers recover without treatment?
8. Why might the ratio of male-to-female stutterers change with age?
9. In what ways is stuttering predictable? In what ways does it vary?
10. Why is it difficult to answer the question, “What is the cause of stuttering?”
11. The International Classification of Functioning, Disability, and Health (ICF) indicates that with some conditions, interpersonal interactions may be affected. How might stuttering affect these interactions?
12. How can stuttering treatment help change factors affecting the individual in theICF area called “Contextual Factors?”
13. How would you describe the etiology of stuttering to a parent who has had limited education and is not used to discussing abstract concepts?
1. Enlist the help of an adult who stutters, and have him teach you to stutter.Then ask him to go with you while you use some voluntary stuttering in public. Write a brief report of what feelings you experienced and how people reacted to you.
2. Use an Internet search engine like Google to find an online discussion groupof people who stutter and clinicians. Join the group and observe what issues they discuss.
3. Attend a support group for people who stutter.
4. Listen to some preschool children talking and note their normal disfluencies. Then listen to elementary school children talking, and compare their disfluencies to those of the preschool children.
5. If you are a fluent speaker, record your own speech and observe the types ofdisfluencies you hear. Do they differ from stuttering? In what ways?
6. Conduct a search on the Internet for resources that guide you to criticallyevaluate Web sites (for example, Using that format, critically evaluate a Web site you find when searching for “stuttering” sites with your search engine.
The King’s Speech. This film depicts King George VI of England who stuttered severely, but found a great deal of help with his Australian speech therapist, Lionel Logue. The movie provides an excellent depiction of the emotions surrounding stuttering.
Bloodstein, O. (1993). Stuttering: The search for a cause and cure. Boston:

Allyn and Bacon.
This book is part history and part analysis, written with charm and clarity. Bloodstein covers early treatments for stuttering, the burgeoning of research in the 1930s, 1940s, and 1950s, and more recent findings in the realm of neurophysiology. His own orientation on the learning-environmental basis of stuttering comes through, but he gives good coverage of other possible factors as well. Bloodstein is particularly good at conveying the excitement that accompanies research.
Bobrick, B. (1995). Knotted tongues–Stuttering in history and the quest for a cure. New York: Simon and Schuster.
A highly readable account of various treatments for stuttering throughout the ages and of famous people who stutter.
Carlisle, J. (1985). Tangled tongue: Living with a stutter. Toronto: University of Toronto Press.
An eloquent autobiography by a man with a severe stutter and a great sense of humor.
Helliesen, G. (2002). Forty years after therapy: One man’s story. Newport News, VA: Apollo Press.
An autobiography of someone who stuttered severely and was treated by Charles Van Riper, the world-renowned stuttering clinician. This book presents a unique view of stuttering therapy from the client’s viewpoint.
Jezer, M. (1997). Stuttering: A life bound up in words. New York: Basic Books.
A compelling, sensitive book about the frustrating and sometimes funny things that happen to someone growing up with a severe stuttering problem and learning to cope with it.
Murray, F.P. (undated). A stutterer’s story. Memphis: Stuttering Foundation of America. (
An autobiography depicting the long struggle of someone who stuttered severely and spent his life searching for answers. The author describes his acquaintance with many of the pioneers of stuttering therapy.
Shields, D. (1989). Dead languages. New York: Knopf.
This is a novel about a young boy who stutters. It conveys the feelings associated with being a stutterer in a world that prizes spoken language. It is recommended for students who would like to understand a child who stutters.
St. Louis, K. (Ed.) (2001). Living with stuttering–Stories, basics, resources, and hope. Morgantown, WV: Populore Publishing Company.
The life stories of 25 people who stutter and how they have coped with their stuttering. Borderline Stuttering
1Yairi and Ambrose (2005) ask whether the age of stuttering onset is getting younger and whether this reflects a general trend of earlier language acquisition. In this regard, Nan Ratner (personal communication, July, 2009) has surveyed a number of language acquisition experts through the Childes discussion group, and their consensus seems to be that there is no clear evidence that the age of language acquisition (i.e., ages at which major language milestones are achieved) has changed over the last few decades. She points out, however, that there is some evidence from differing results achieved by older versus more recent studies of phonological development that children may be acquiring articulatory targets earlier. Whether this would affect the age of stuttering onset is not known.
Constitutional Factors in Stuttering

Biological Background
Hereditary Factors
Family Studies
Twin Studies
Adoption Studies
Congenital and Early Childhood Factors
Brain Structure and Function
Brain Structure Differences in People Who Stutter
Brain Function Differences in People Who Stutter Sensory and Sensory-Motor Factors
Some Caveats about Finding the Causes of Stuttering
Sensory Processing
Central Auditory Processing
Brain Electrical Potentials Reflecting Auditory Processing
Dichotic Listening Tests Auditory Feedback
Other Sensory Feedback
Sensory-Motor Control
Reaction Time
Fluent Speech
Nonspeech Motor Control
Language Factors
Emotional Factors
Anxiety and Autonomic Arousal
A Lay Description of Constitutional Factors
After studying this chapter, readers should be able to:
• Describe the evidence supporting genetic inheritance of stuttering from (a) family studies, (b) twin studies, and (c) adoption studies
• Explain how genetic studies are done and what has been found about howchromosomes and genes are associated with stuttering
• Explain why congenital and early childhood factors are thought to contribute toa predisposition to develop stuttering and what some of these factors may be
• Describe the ways in which brain structures in stuttering individuals differ fromthose of nonstuttering individuals
• Describe the ways in which brain functions in stuttering individuals differ fromthose of nonstuttering individuals
• Describe changes in the brain that are associated with improvements in fluencyas the result of treatment
• Describe differences in these areas found between groups of individuals whostutter and groups of individuals who don’t: (a) sensory processing, (b) central auditory processing, (c) dichotic listening, (d) auditory feedback, (e) sensorymotor control, (f) reaction time, (g) fluent speech, and (h) nonspeech motor control
• Suggest why differences in each of the above areas could contribute to stuttering
• Describe how language development and performance in individuals who stutter have been found to differ from individuals who don’t
• Describe the ways in which stuttering and emotion may be related.
Anomaly: A difference from the normal structure or function
Family studies: Examination of family trees of individuals who stutter to determine the frequency and pattern of the occurrence of stuttering in relatives. These studies can answer questions such as whether males or females are more likely to have children who stutter and whether persistent stuttering (as opposed to natural recovery) is a trait that is inherited
Twin studies: Research on the co-occurrence of stuttering of both members of a twin pair if one twin stutters; questions such as whether identical twins show more concordance than fraternal twins can be answered, shedding light on the extent of a genetic basis of stuttering
Adoption studies: Investigations of stuttering in siblings who were adopted soon after birth and placed with different families. A higher incidence of stuttering among biological relatives than adoptive family members also provides evidence of a genetic basis of stuttering rather than an environmental basis Persistent stuttering: Stuttering that persists for several years after onset, beyond the time at which natural recovery is likely to occur
Natural recovery (from stuttering): Stuttering that disappears within a year or two after onset from natural causes rather than from treatment
Concordance (in twins): If one twin has a condition, such as stuttering, the other twin also has the condition
Gene: A segment of DNA that determines an individual’s traits, such as height and weight
DNA: A double-stranded molecule passed on from a mother and a father to a child containing the “instruction book” for passing on traits
Congenital factor: A physical or psychological trauma that occurred at or near birth that may predispose an individual to develop stuttering
Predisposition: A susceptibility to developing a condition
Sensory processing: Activity of the brain as it interprets information coming from the senses, such as sounds arriving via the ears and auditory nerves
Sensory-motor control: The way all movement is carried out with sensory information used before, during, and after to improve the precision of movement
Temperament: Aspects of an individual’s personality, such as sensitive versus thick skinned, that are thought to be innate rather than learned
I have written this chapter to be a readable summary of perspectives and research on constitutional factors in stuttering. Details of important studies are provided in Chapter 3. Figure 2.1 illustrates the major areas covered in this chapter and how the topics are related.

Figure 2.1 Constitutional factors in stuttering.
Taking it from the top, you will learn that heredity and congenital and/or early childhood trauma are background factors that can contribute to stuttering. They influence how brain structure and function develop. You will also learn how the brains of those of us who stutter may be different from those of typical speakers and how that may influence what I believe is the major direct link to stuttering–sensory-motor control. You will also learn about two other factors that appear to be influences on stuttering onset and development via their influences on sensory-motor control of speech: language factors and emotional factors. Both of these are to some extent products of brain structure and function. Language and emotional factors are also influenced by the environment, a major determinant of stuttering that will be covered in Chapters 4 and 5.
Hereditary Factors
Two individuals contributed a vast amount to our understanding of the mechanisms of heredity. One was the Austrian monk, Gregor Mendel, whose experiments in breeding varieties of peas gave rise to his insights about genetic inheritance. He established the principle that each parent in a breeding pair contributes equally to the genetic makeup of the offspring, and he developed the understanding of dominant and recessive traits. The other major contributor to the science of genetics was Charles Darwin, who, interestingly, is thought to have inherited stuttering from his grandfather, Erasmus (Thomson, 2009). Darwin’s theory of evolution suggested that while specific characteristics (such as height, weight, and skin color) could be inherited by one generation from another, variations in how those characteristics expressed themselves produced slight differences in members of a species. These differences favored some individuals more than others, depending on the environment; the most favorable traits for the current environment of the species would be increased because individuals possessing them would thrive and reproduce most frequently.
Stuttering often runs in families, a fact long recognized by researchers (e.g., Andrews et al., 1983; Bloodstein & Ratner, 2008). For many years, researchers debated about what this meant. Some suggested that the appearance of stuttering in several generations of a family must mean that it is caused by an inherited neurological difference or anomaly. Others disagreed, countering that political beliefs often run in families too, but aren’t inherited. Some researchers argued that stuttering develops in response to a critical attitude toward normal disfluency that has been handed down from one generation to the next (Johnson & associates, 1959). A child whose parents were critical of her normal disfluencies would grow afraid and would “hesitate to hesitate.” This would start a spiral of more hesitations leading to greater fear and so on.
For many years, researchers aligned themselves with one side of this argument or the other. Currently, however, there is broad agreement that stuttering can be inherited (Bloodstein & Ratner, 2008; Yairi & Ambrose, 2005). In other words, for many people who stutter, one or both of their parents had some predisposition to stuttering that was transmitted in their genes. Current thinking may be due to strong newer evidence about heredity in stuttering but is probably also due to the rise of less deterministic views of heredity. Research has shown, for a number of inherited disorders, that genes do not work alone. Stuttering, asthma, migraine headaches, and certain other disorders are seen as the result of heredity and environment acting together with elements of chance thrown in (Kidd, 1984). The interaction between heredity and environment is something we commonly encounter. For example, this year I grew all my tomato plants from the same seed packet, giving them an identical heredity. But I planted some near Burlington, in a relatively warm environment (for Vermont), and I planted others in the Northeast Kingdom of Vermont, a colder, cloudier summer environment. As you might guess, by mid-August the tomatoes near warmer, sunnier Burlington were plump and red, while those in the Northeast Kingdom were shriveled and sickly green, showing that environment had a strong differential effect on heredity. And so it is with stuttering.
Going back to the topic at hand, a child in one family may inherit genes predisposing him to stutter, but his home environment may be so low key that stuttering never develops. A different child, inheriting similar genes, may grow up in a demanding, hectic, fast-talking home and begin to stutter at age 3. And lest this comment appear to hearken back to a period when parents were blamed, it may often be the case that which aspects of the environment play a role for any given child will remain a mystery.
In this chapter, I will review three approaches to the study of heredity and stuttering: family studies, twin studies, and adoption studies. These different ways of gathering evidence all suggest that for many individuals, stuttering is partly attributable to heredity. The insights we gain from these studies are vital to us in counseling individuals and families about the nature of stuttering.
Family Studies
In family studies, researchers gather evidence of the inheritance of stuttering by studying family trees. They begin with a group of individuals who stutter, for example, a group of 100. They then make up a group of 100 individuals who don’t stutter, matched with the first group by age and gender. The researchers interview all the relatives in each group member’s family to determine the average number of stuttering relatives each person has. Then they compare the findings in each group to answer the question: Do individuals who stutter have more relatives who stutter than individuals who don’t stutter?
As scientists study family trees of stutterers, they also search for patterns of occurrence of stuttering. Geneticists know that certain inherited traits occur in specific patterns in families, and they ask whether stuttering follows a known pattern of inheritance. For example, some traits appear only if both the mother and father have the trait; other traits are more common in children when the mother has the trait than when the father has the trait. Therefore, when scientists find these known patterns of family occurrence in a disorder like stuttering, it provides more evidence to support the idea that the disorder is inherited rather than simply the result of imitation, critical family attitudes about disfluency, or some other aspect of their shared environment.
To summarize the research, despite the small number of studies and their limitations, family studies have provided strong evidence for a genetic predisposition in many individuals who stutter. Males tend to be at more risk to develop persistent stuttering, while females seem to have some natural resistance to persistent stuttering and tend to recover from stuttering more easily. Some evidence also suggests that there may be more than one genetic mechanism involved in persistent stuttering; one or more genes may carry the predisposition for the speech breakdown evident in children who begin to stutter, while an additional genetic predisposition may prevent or facilitate natural recovery from stuttering.
Twin Studies
The genetic transmission of stuttering can also be investigated by comparing the incidence of stuttering in fraternal and identical twins. Identical twins (also called monozygotic twins) have completely identical genes. In fraternal twins (dizygotic), only 25 percent of their genes are identical, like any other two siblings. Greater similarities in the traits of identical twins compared with those of fraternal twins are generally attributed to inheritance.
The many studies of stuttering in twins reveal two major findings. One is that compared to fraternal, same-sex twins, identical twins show more concordance (if one twin stutters, the other also does). The second is that even though there is much concordance among identical twins, there are still many identical twin pairs in which one twin stutters and the other doesn’t (discordance). These two findings suggest that genes don’t work alone; the environment must interact with genes to produce the behavior in question, and even in twins, their environments (both before they are born and after) may not be as similar as they might superficially seem.
Adoption Studies
One of the most powerful methods to examine the relative contributions of genes and the environment to stuttering is to look at the families of stutterers who were adopted soon after birth. Higher incidence of stuttering among the biological relatives of adopted stutterers would support a greater role of genetics in causing stuttering, whereas higher incidence among adoptive relatives would support a greater role of the environment. The few studies that have been done had mixed results, but one study that investigated both biological and adoptive families found that although both heredity and environment play a role in the occurrence of stuttering, heredity plays a slightly stronger role (Felsenfeld, 1997).
Humans have between 25,000 and 35,000 different genes. These genes are segments of DNA that determine various individual traits. DNA contains the “instruction book” that tells the body how to make various chemicals that determine such characteristics as a person’s height, weight, eye color, emotional vulnerability, athletic ability, and everything else that is influenced by heredity. Very long strings of DNA (containing many genes) are wrapped into the little worm-like structures called chromosomes (see Fig. 2.2).

Figure 2.2 Chromosomes and gene.
Almost every cell in the body contains 23 pairs of chromosomes–one member of each pair from the individual’s mother and one from the father. Some traits may be determined more by the mother’s genes and others more by the father’s. Some traits are determined by one gene, others by multiple genes acting together. The hunt for genes related to stuttering is conducted by looking at the chromosomes in many related individuals, some of whom stutter and some of whom don’t. Thus, many chromosomes in related individuals will all be the same, but if genetics is a factor in stuttering, the subgroup of related people who stutter will have one or more chromosomes that will be different from the subgroup of related people who don’t stutter.
Researchers looking at families and communities with large concentrations of individuals who stutter have identified seven different chromosomes that appear to be associated with stuttering. One of these chromosomes has been shown to carry three genetic mutations, one of which is associated with motor control and emotional regulation (Kang, Riazuddin, Mundorff, Krasnewich, & Friedman et al., 2010). The many different chromosomes and genes found to be associated with stuttering in the studies described in Chapter 3 testify to how varied and complex the origins of stuttering may be. Many different factors may contribute to stuttering, and these factors interact with others to create a neurological substrate for stuttering. For example, neurophysiological deficits may give rise to problems and delays in speech motor control, sensory processing, language planning and execution, and/or emotional vulnerability. Combinations of these problems may result in the repetitions, prolongations, and blocks that appear in children as they develop speech and language in “one great blooming, buzzing confusion” (James, 1890) that is the child’s normal environment in the typical family.
When talking with parents about the genetics of stuttering, be aware that parents may have mixed feelings about the possibility that their child inherited a predisposition to stutter. On the one hand, it may help them realize that their parenting style didn’t cause the stuttering; on the other, parents may feel guilty for passing on genes that may have resulted in stuttering. However, it is important for them to know that genetic variations are neither all good nor all bad. Remember the genius of Charles Darwin: Genes that predispose a child to stuttering may also pass on very beneficial or even extraordinary talents (Table 2.1).
Table 2.1 Summary of Hereditary Factors in Stuttering

Congenital and Early Childhood Factors
It has been estimated that 30 to 60 percent of individuals who stutter have family histories of stuttering (Yairi, Ambrose, & Cox, 1996). Therefore, 40 to 70 percent of individuals who stutter have no family history and may have developed the disorder by mechanisms other than inheritance of a predisposition that has caused stuttering in relatives. However, we do not know how many of these individuals inherited a factor or several factors that predisposed them to stuttering but did not produce stuttering in other family members. On the other hand, some individuals who stutter may not have inherited any factor, which predisposed them to stutter. Instead, they may have experienced a physical or psychological trauma that predisposed them to stuttering or even precipitated its onset. Such traumas may have occurred at or near birth and would be viewed as congenital factors. Some events or factors may also have occurred during early childhood rather than at or near birth, but for simplicity’s sake, I will refer to them all in the remainder of this section as congenital.
Unfortunately, there is relatively little research on congenital factors related to stuttering. The research that has been done can be divided into studies which examined stutterers who had no family history of stuttering and studies that examined large groups of individuals who sustained brain injuries at birth or in childhood to determine if there were more cases of stuttering in populations with early brain injuries than in the general population.
In summary, studies of individuals who stuttered but had no family history of stuttering found that this group had more history of infectious diseases, anoxia at birth, childhood surgery, head injury, mild cerebral palsy, mild retardation, and intense fear prior to the onset of stuttering. Studies of young adults who had brain injuries found a greater incidence of stuttering, especially if the individuals had been unconscious for a time following the injury. Finally, although the results are not uniform in this, studies of stutterers with family histories of stuttering found that these individuals had more prolongations and silent blocks in stuttering and a slower rate and more variability in fluent speech, compared with stutterers without a family history of stuttering.
As we’ve just seen, the predisposition for stuttering may come not only from genetic inheritance but may also come from congenital factors. Thus, it may be most appropriate to use the term “constitutional predisposition” (Table 2.2).
Table 2.2 Summary of Congenital and Early Childhood Factors

Brain Structure and Function
Whether an individual’s stuttering results from an inherited predisposition or from an early brain injury or other trauma, it seems likely that structures and functions in the central nervous system would be different or “anomalous” in those who stutter. Brain structures and functions would be, in effect, the bridge between the etiology of the disorder and the behavior. In other words, a genetic predisposition, injury, trauma, or some unknown cause might result in brain structures and/or functions that delay or disrupt normal neural processing for speech and language. This delay or disruption might, in turn, result in the repetitions, prolongations, and/or blocks that characterize the disorder.
The search for how these brain structures or functions are different in people who stutter has been underway at least since the experiments of Lee Edward Travis at the University of Iowa in the 1930s, described in Chapter 1. In the following sections, I will review studies of brain structure and brain function in separate sections, even though they obviously influence each other. I will follow the discussion of structures and functions by talking about how stuttering may have changed brain structure and how treatment may have changed brain function. I am indebted to the chapter by Neumann and Euler (2010) for some of the recent brain imaging information. As you read about specific anatomical areas, consult Figure 2.3 to orient yourself to landmarks related to speech and language production in both stutterers and nonstutterers.

Figure 2.3 Areas in the left and right sides of the brain that may be involved in speech and language processing in stutterers and nonstutterers.
Brain Structure Differences in People Who Stutter
Compared to the search for anomalous function in the brains of people who stutter, research on actual structural differences in their brains has begun relatively recently. Several studies between 2000 and 2007 (see Chapter 3) examined the brain anatomy of adults who stuttered by measuring the shape, size, and density of speech and language areas. The findings suggest that sensory, planning, and motor areas in the left hemisphere of these individuals developed differently from those in matched nonstuttering individuals. For example, white matter tracts, which convey information from sensory centers that may store phonological representations of sounds to motor execution areas of the left hemisphere, have been shown to be less dense than those in normal speakers, whereas the same tracts were found to be denser in the right hemisphere of those who stuttered. It is as if the right hemisphere had taken over some typically left hemisphere functions.
By 2008, neuroimaging researchers had developed techniques that were safe enough to use with children. In that year, two groups of investigators published studies of school-age children who had stuttered in their preschool years. One study of children who recovered from stuttering compared to those who hadn’t (and a control group) showed that both recovered and persistent stutterers had reduced volumes of gray matter around Broca’s area, as well as in bilateral temporal lobe areas that may be related to auditory perception of speech (Chang, Erickson, Ambrose, HasegawaJohnson, & Ludlow, 2008). The subgroup that persisted in stuttering also showed less dense white matter tracts connecting phonological representations of sounds to speech motor execution areas, the same deficit as discovered in adult stutterers, described earlier. This finding was reported again in a second study of slightly older children in that same year, suggesting that this structural abnormality in the left hemisphere may be a major factor in stuttering (Watkins, Smith, Davis, & Howell, 2008).
The 2008 findings of the two groups cited above were replicated by Cykowski, Fox, Ingham, Ingham, and Robin (2010), using more extensive brain imaging technology. They suggested that the most robust difference between stutterers and nonstutterers is in left-hemisphere fiber tracts that communicate between the inferior parietal cortex (sensory integration) with the ventral frontal cortex (motor planning) (see Fig. 2.4). As in earlier studies, the authors found that in individuals who stutter compared to nonstutterers, certain nerve fibers aren’t structured as effectively to conduct impulses along the directional flow of the nerve bundle. Thus, conduction is not as fast as it might be. The fiber tract in question is the superior longitudinal fasciculus; its function is to provide sensory-motor integration for speech. See the next chapter for more details on these research findings.

Figure 2.4 The superior longitudinal fasciculus III is a bi-direction pathway between the inferior parietal cortex (sensory integration) and the ventral frontal cortex (motor planning).
Brain Function Differences in People Who Stutter
Research on functional differences has a long history beginning in the 1920s with what now seems like primitive technology. Interestingly, more recent studies using modern technology have validated many of those old findings.
In essence, both old and new studies have shown that individuals who stutter have a greater activity in their right hemispheres than in their left hemispheres, during both fluent and stuttered speech. This is the reverse of the pattern shown by fluent control subjects who show considerable left-hemisphere activity and little right-hemisphere activity during speech. Figure 2.5 shows this difference between stutterers and nonstutterers. The activity seen in the stutterers’ right hemispheres was often in those very areas that are homologous (parallel) to the left-hemisphere areas most active in fluent speakers.

Figure 2.5 PET scans of brains of nonstuttering (left) and stuttering (right) adults while reading aloud. SPM, statistical parametric mapping. (From De Nil, L., Kroll, R., Kapur, S., & Houle, S. (1995). Silent and oral reading in stuttering and nonstuttering adults: A positron emission tomography study. Paper presented at the Annual Convention of the American Speech-Language-Hearing Association, Orlando, Florida, December.)
Neuroimaging studies have also revealed a great deal of underactivation of lefthemisphere structures typically active for speech, such as areas around the white matter tracts presumed to carry information from sensory and sound representation areas to motor execution areas (see Figs. 2.4 and 2.6).

Figure 2.6 Density of white matter tracts between sensorimotor areas and speech planning areas in stutterers (study group) and nonstutterers (control group).
Studies of stutterers after treatment have generally revealed a reversal of righthemisphere overactivation and left-hemisphere underactivation just described. Both short-term and long-term treatment outcome studies using brain imaging suggested that areas of the left-hemisphere that were previously underactivated were now reactivated, and right-hemisphere sites were now more normally underactivated.
Figure 2.7 depicts the differences in brain activity between stutterers before and after treatment. It is evident, particularly in the brain images of stutterers in Figure 2.7 (a) before treatment and (b) immediately after treatment, that activity levels shift from greater in the right hemisphere to greater in the left hemisphere (Table 2.3).

Figure 2.7 Overt reading: statistical parametrical maps of between-group comparisons (people who stutter versus people who do not stutter) (A) before therapy, (B) immediately after therapy, (C) two years after therapy (Neumann, Euler, Wolff von Gudenberg, Giraud, & Lanfermann et al., 2003).
Table 2.3 Summary of Brain Structure and Function in Stuttering

As you can see in Figure 2.1, sensory and motor factors that influence stuttering emerge from an individual’s brain structure and function. They influence the onset and development of stuttering because they limit how well the individual can produce rapid and fluent speech. For example, earlier I described the findings that suggest that the fiber tracts integrating sensory, planning, and motor functions for speech appear to be significantly less dense in individuals who stutter. This may mean that sensorymotor tasks, such as saying “ah” after hearing a bell, will be slower in stutterers.
The findings summarized in the following sections compare stuttering and nonstuttering speakers’ sensory and sensory-motor functions, using a variety of experiments. Keep in mind that the purpose of these comparisons is to learn more about deficits in those who stutter. When sensory and sensory-motor functions are poorer in stutterers, there is an assumption that it reflects a brain structure or function difference that may be causally related to stuttering.
Some Caveats about Finding the Causes of Stuttering
As researchers have searched for the causes of stuttering, they have been hampered by the difficulty in demonstrating cause-and-effect relationships. In particular, the brain differences described in the preceding section could be the cause of the disorder or the result of how the brain has responded to years of stuttering (most likely, some of each is involved). Moreover, brain activity during stuttering may also be indirectly related to the behavior of stuttering; for example, some right hemisphere activity may be a manifestation of stutterers’ fears and anxiety about stuttering. These fears might not cause the disorder in the first place, but they may make it worse.
In some speech disorders, which have a more clear-cut physical basis, cause and effect can be directly investigated. There may even be animal models of the disorder to work from. In a disorder that can occur in animals, factors thought to cause the disorder can be manipulated and their effects measured. For example, interfering with the genes or the embryonic development of dogs, cats, or rabbits can produce the clefting or spasticity that is respectively similar to cleft palate or cerebral palsy seen in humans. Thus, scientists can infer how inheritance, embryonic development, or brain trauma can cause these disorders. However, communication through spoken language does not occur in animals, which makes it less likely that animal models can be used to look for the cause of disorders like stuttering. And obviously, selective breeding or surgery to create stuttering is not an option in humans. Therefore, researchers have turned to indirect approaches, or descriptive rather than experimental approaches. They compare groups of stutterers and nonstutterers on tasks they believe are related to speech fluency. If they find, again and again, that the two groups perform differently in certain tasks, they may have a clue about the disorder.
Such indirect research is complicated because the differences that are found might be a result of stuttering, not a cause of it. For example, reaction time studies of how quickly subjects can say a word that is flashed on a screen might show that subjects who stutter are slower than subjects who don’t. This difference, however, might be the result of subjects who stutter saying words more slowly to keep from stuttering. Even if a difference is not the result of trying not to stutter, it is more likely to be only a correlated factor that has no causal relationship to stuttering. Even if groups of stutterers do respond more slowly, slower reaction times alone probably do not cause stuttering. If they did, people would start stuttering as they grew older or after they had imbibed a few beers. Finding something that co-occurs with something else but doesn’t cause it is like finding that most basketball players have larger shoe sizes than most gymnasts. Shoe size itself doesn’t determine who is better suited for each of these sports, but height may be a determining factor. Both shoe size and height are related to one’s genetically determined bone size, so they end up being different in gymnasts and basketball players, but shoe size itself doesn’t make one a better basketball player or gymnast.
Another problem with descriptive studies rather than experimental approaches to studying the nature of stuttering is that when comparing groups of people who do and do not stutter, there is often a great deal of overlap in the two groups’ performances even though their averages might be statistically different. For example, some of the people who stutter will usually show coordination as good as the average person in the group of nonstutterers on a typical test of motor coordination, and some of the people who don’t stutter will demonstrate the same level of coordination as an average person in the stuttering group. These overlaps remind us that we are usually not studying factors that are necessary or sufficient by themselves to cause stuttering. Such underlying differences between people who stutter and people who don’t may provide us clues, however. And with those clues in hand, we may look more closely at certain abilities, brain functions, and neuroanatomical sites to see if there truly are things that distinguish all stutterers from all nonstutterers. Failing to find that, we will see if there are subgroups of stutterers, some of whom differ from nonstutterers in one way and others who differ in another way. This may then help us discover other paths leading to a better understanding of stuttering.
In the next section, I will review some of the literature on stutterer/nonstutterer performance differences. As you will see, when scientists try to repeat others’ experiments to verify their results, inconsistent research findings are common. One study finds a difference; another study reports there isn’t one. Differences in the findings of two studies may occur because different subjects were involved or because there were small differences in the way that the studies were done. For example, one study may use a 1,000-Hz tone as a stimulus, and another study may use a recording of the word “go.” Despite the inconsistencies in many results, there are areas of agreement or trends that many studies find. As you read, try to determine for yourself which areas give us solid leads. In my summaries, I will share my own interpretation of these areas of overlap.
Sensory Processing
You might wonder why research has been conducted on the ability of individuals who stutter to process sensory information such as auditory, visual, and tactile signals. Stuttering, after all, appears to be a motor rather than a sensory problem. The answer is twofold. First, as patients with various injuries and diseases have taught us, normal speech depends on intact auditory as well as proprioceptive (feeling of position and movement) and tactile (feeling of touch) feedback. Researchers have been curious to see if stutterers’ abnormal speech might be the result of some disturbance of feedback. Second, experiments, which have altered sensory processing, such as delayed auditory feedback (Black, 1951; Lee, 1951), have created repetitions, prolongations, and blocks in normal speakers, prompting scientists to ask whether this might be the cause of stuttering.
The findings that I’m about to review may be related to the results of the brain imaging studies discussed earlier. Remember that many studies found that areas of the auditory cortex are underactivated during stuttering (Beal, Gracco, Lafaille, & De Nil, 2007; Brown, Ingham, Ingham, Laird, & Fox, 2005), while others have found structural anomalies in the auditory cortex (Foundas, Bollich, Corey, Hurley, & Heilman, 2001). Still other research has discovered reduced density in stutterers’ white matter fibers that support sensory-motor integration in speech production (Sommer, Koch, Paulus, Willer, & Bücher, 2002). It would not be surprising, then, if anomalies in these areas affect both the fluency of speech production and the accuracy of speech perception. For example, the superior temporal gyrus has been shown to contain systems that are important in the phonemic planning of utterances and the understanding of speech (Hickok, 2001). Moreover, efficient functioning of the auditory cortex is likely to be critical for fluent speech production because of the crucial role of auditory feedback in normal speaking and the deleterious effects on fluency of delaying auditory feedback.
Central Auditory Processing
A number of studies (see Chapter 3) have been conducted to assess how accurately and quickly individuals who stutter can identify or judge the duration of auditory signals, compared to nonstutterers. The major findings have been that stutterers are poorer at processing auditory signals. They are less accurate at identifying words and sentences in noisy conditions, and they are poorer at judging the durations of tones. Interestingly, researchers have also found that nonstutterers who are more disfluent perform more poorly at the tests than do nonstutterers who are more fluent. Remember the findings that individuals who stutter have deficits or anomalies in auditory processing areas of the brain? It seems likely that these deficits are responsible for the poorer scores on tests of central auditory processing.
Brain Electrical Potentials Reflecting Auditory Processing
Although there are conflicting findings when experimenters have looked at brain waves associated with listening to auditory stimuli, there appears to be a subgroup of stutterers who have anomalous responses. This subgroup showed longer delays between the stimuli and the brain wave responses, as well as smaller brain waves. This finding may reflect the findings in brain anatomy studies that identified a subgroup with anomalies in the auditory cortex that benefit more than others from delayed auditory feedback to promote fluency. These and other similar findings will be discussed further in Chapter 3.
Dichotic Listening Tests
In the early 1960s, a procedure was developed to assess hemispheric dominance for speech and language by testing which ear was more accurate at hearing speech sounds. Kimura (1961), a Canadian psychologist, invented the “dichotic listening test,” which simultaneously presented two different syllables (like “ba” and “da”) dichotically–a different syllable to each ear. Listeners reported which syllable they heard. Auditory nerves connecting the ears to the cerebral hemispheres carry more information to the hemisphere on the opposite side than to the hemisphere on the same side. Results with normal speakers indicated that syllables presented to the right ear (opposite the left hemisphere, which is dominant for speech and language) were most frequently reported as heard, which was called a right ear advantage for speech.
This procedure has been used to assess laterality differences between stuttering and nonstuttering groups. Again there have been conflicting findings, but most of the dichotic tests that have used linguistic stimuli such as words and sentences have found that individuals who stutter have reversed hemispheric dominance for perception of speech. That is, nonstutterers tend to have left-hemisphere dominance, and stutterers have more right-hemisphere dominance. As with other studies of auditory processing, researchers are beginning to suspect that not all stutterers are alike in this respect; there may be a subgroup that has anomalous auditory processing.
Auditory Feedback
Ever since the ancient Greek stutterer Demosthenes improved his speech by orating above the roar of the Mediterranean Sea, it has been observed that changes in auditory feedback can affect fluency. Masking noise, delayed auditory feedback, frequency shifts, and other alterations in the properties of the auditory signal can create temporary fluency in persons who stutter (see Van Riper, 1982, for a review up to that date). On the other hand, delayed auditory feedback can create an artificial stutter in normal speakers (Black, 1951; Lee, 1951). A variety of explanations for the effects of altered feedback on people who stutter have been offered, including that it (1) is a distraction, (2) causes stutterers to change how they talk (e.g., becoming louder), and (3) compensates for a defect in stutterers’ auditory monitoring of their speech (Bloodstein, 1995; Garber & Martin, 1977).
Other Sensory Feedback
The findings reviewed in the last few sections are related to the auditory system, but other sensory systems are important for the control of speech–specifically, touch and movement. A number of studies have suggested that stutterers have poorer sensory feedback in several domains, but other studies don’t find these differences. Perhaps that is why some of the sensory feedback enhancements used for treatment (e.g., SpeechEasy) are effective with some individuals who stutter but not others (Foundas, Bollich, Feldman, Corey, & Hurley et al., 2004; Pollard, Ellis, Finan, & Ramig, 2009) (Table 2.4).
Table 2.4 Summary of Sensory Processing and Stuttering

On the face of it, stuttering appears to be a disorder of speech motor control. Fluent speech depends on sensory-motor control of the muscles that move speech structures to produce airflow, voicing, and articulation in a coordinated fashion so that speech sounds are produced smoothly, in a specified sequence, and at a reasonable rate. Stuttered speech, then, must be the result of some disturbance in the smooth, sequenced muscle contractions necessary for coordinated structural movements. Van Riper described this as “a temporal disruption of the simultaneous and successive programming of muscular movements” (Van Riper, 1971, p. 404; 1982, p. 415).
The control of the smooth movements of speech depends in part on sensory input as well as motor output. In fact, part of the control of any complex movement uses sensory information about where the structure is now and where it’s going in order to produce just the right amount of contraction of all the muscles involved. When the brain plans the movements needed to produce sounds, it uses stored memories of past movements and their consequences in planning what must be moved as well as when and how to produce the desired acoustic and perceptual result–the sounds of speech. This section reviews several areas of research that have looked at stutterers’ speech and nonspeech motor control. It’s important to remember that even if it is not explicitly stated by researchers, investigations of motor skills are in fact investigations of sensory-motor skills.
Reaction Time
Figure 2.8 depicts an example of a reaction time experiment. The participant is told to watch the computer screen for a picture of an object and to say the name of the object the instant it appears. The time between the appearance of the object on the screen and the first sound or movement made by the participant is her reaction time. As indicated, reaction time involves sensory analysis, response planning, and response execution. It is therefore a potentially useful measure in stuttering research if it is thought that the core deficit is a delay in some aspect of sensory processing, planning, or motor execution.

Figure 2.8 Processing stages in a reaction time task.
Beginning in 1976, experimenters have shown that individuals who stutter
(including children) often have slower reaction times than individuals who don’t (see Chapter 3 for detailed descriptions of these studies). This has been shown with both auditory and visual stimuli and with responses involving initiating and terminating a vowel sound, pressing lips together, and making respiratory movements. Differences were more frequently found when linguistically meaningful stimuli were used to test reaction time. These differences probably reflect the brain imaging evidence discussed earlier in this chapter that individuals who stutter have anomalies in brain areas related to sensory-motor integration.
Fluent Speech
Reaction time responses, such as lip pressing or saying “ahhh,” are indirect measures of speech processing under normal conditions. Researchers have been able to make more direct assessments by examining the speed and coordination of stutterers’ speech movements when they are talking fluently and by analyzing the sound waves of their fluent speech. The findings generally show that people who stutter have slower speech movements and sometimes have abnormal sequencing in the movements of their articulators (see Chapter 3). There are different interpretations of these findings, including views that (a) stutterers have sensory-motor delays caused by abnormalities in brain pathways, (b) these findings only reflect strategies that stutterers use to be fluent, and (c) stutterers are slower in their speech production because, even in some of their apparently fluent speech, they abnormally tense speech muscles in a way that puts a drag on the movement of their articulators. I suspect that each of these three interpretations may be true for different subjects in different experiments.
Nonspeech Motor Control
Following in the wake of studies that showed individuals who stutter often have slower than normal reaction times and slower segments in their fluent speech, researchers began to examine complex motor coordination of nonspeech muscles and structures. One advantage of this approach is that it eliminates the effect of stuttering itself that may contaminate measures of speech movements. Second, complex motor coordination, such as sequential finger movements, appears to be planned and organized by areas of the brain, such as the supplementary motor area (SMA), which also appear to be involved in the sequential articulatory movements of speech (Goldberg, 1985). Interestingly, neurophysiological evidence that sequential finger movements are regulated by the same brain regions that control speech has been supported by arguments that spoken language in humans evolved from right-hand (and thus left-hemisphere) specialization for manipulating objects and hand gesturing in earlier hominids (Kent, 1997; MacNeilage, 1987).
A number of studies have shown that in various nonspeech tasks such as tapping with fingers in a prescribed sequence, stutterers are slower than nonstutterers. They may be significantly slower in initiating the sequence of taps, and more severe stutterers may be notably slower than mild stutterers or nonstutterers. Finger tapping experiments have also suggested that right-hemisphere activity may interfere with stutterers’ dominant (right)-hand sequential finger tapping, which requires input from the left hemisphere. In fact, other experiments showed that both right-handed stutterers and left-handed non-stutterers are less able to focus on one hemisphere without interference by the other. Another interesting study of nonspeech motor control suggested that stutterers’ optimal tapping rate was slower than that of nonstutterers; however, when instructed to tap at a fast rate, stutterers could tap very fast but became highly unstable (high degree of variability) when doing so (Subramanian & Yairi, 2006). This is in agreement with a study of children’s hand clapping, which demonstrated that a subgroup of children who stuttered was highly variable compared to children who didn’t stutter when clapping their hands at a specified rate (Olander, Smith, & Zeleznik, 2010). Researchers looking at adult stutterers’ ability to follow auditory pitch changes using various motor outputs also found the stutterers to be slower and less accurate than nonstutterers.
These experiments suggest that stutterers are slower and/or more variable than nonstutterers at nonspeech motor tasks. Some of these tasks showed that stutterers are poorer at integrating sensory and motor information to produce a movement. Other tasks showed that when they were performing sensory-motor tasks, they lacked the left-hemisphere dominance shown by nonstutterers. The relationship of these findings to stuttering is that many, if not all, individuals who stutter have a basic deficit in sensory-motor control, whether for speech or for nonspeech motor activities (Table 2.5).
Table 2.5 Summary of Sensory-Motor Control and Stuttering

Figure 2.1 depicts language factors as having their influence by affecting the primary deficit in stuttering, sensory-motor control of speech. The influence of language on stuttering is threefold–language development, language delays, and language complexity.
The first factor, language development, exerts its influence because of the stress that language development in a child puts on his speech production. The next chapter considers this factor in detail. But to put it in a nutshell here, the rapid language acquisition that occurs in all children between the ages of 2 and 5 places high demands on brain resources. Add this stress to a child’s basic deficit in sensory-motor control of speech, and stuttering may emerge. As evidence for this hypothesis that children predisposed to stuttering are stressed by language development, researchers point to the fact that the stuttering usually begins at the very time when language growth is greatest (Bloodstein & Ratner, 2008). In fact, Yairi and Ambrose (2005) reported that in more than 50 percent of their sample of stuttering children, parents reported the onset of stuttering during a sudden increase in language development.
The second factor, language disorder or delay, may precipitate or worsen children’s stuttering because these children essentially have two deficits to deal with–a speech motor control problem and a language problem. Again, this may cause children who stutter to divert resources or attention away from compensating for the speech motor control problem to deal with the language problem. Many studies have demonstrated language problems in some children who stutter, but the best recent compendium of these findings is a meta-analysis of 22 studies that compared language ability in samples of stuttering and nonstuttering children (Ntourou, Conture, & Lipsey, 2011). The authors found that four language measures showed significant differences between these two groups: overall language, receptive vocabulary, expressive vocabulary, and mean length of utterance. They speculate that the effect of these language difficulties is that “…when planning/formulating sentences, [children who stutter] may experience subtle, but important difficulties in quickly and efficiently encoding and retrieving lexical items. The end product of such difficulties, it may be argued, is the disruption in the fluent initiation and/or continuation of speechlanguage, most typically characterized by production of speech disfluencies (e.g., revisions, hesitations)” (Ntourou, Conture, & Lipsey, 2011, pp. 174–175).
The third influence of language on stuttering, language complexity, appears to affect sensory-motor control of speech to trigger the occurrence of stuttering in a moment-to-moment fashion. Researchers have found that when individuals who stutter produce longer or more linguistically complex utterances, they are more likely to stutter, whether they are adults or children. Studies have also shown that certain grammatical word types (e.g., nouns and verbs), longer words, and words at the beginning of an utterance are more likely to be stuttered (see Chapter 1 for details). Again, it is as if the extra demand of processing language–in an individual perhaps with other factors such as a genetic predisposition for stuttering–impedes the individual’s sensory-motor control of speech. As I will suggest in the next section, emotion is another brain activity that may disrupt speech production in these children with vulnerable systems. Details on some of these key studies are given in Chapter 3 (Table 2.6).
Table 2.6 Summary of Language Factors in Stuttering

Because people who stutter are a heterogeneous group, the relationship between emotion and stuttering, like the relationship between language and stuttering, will vary among individuals. For some, emotion may be an important etiological factor that triggers the onset of stuttering and makes recovery difficult. For others, although emotion in strong doses may make stuttering change–sometimes for the worse, sometimes for the better–emotion may not be a major factor in its etiology. However, the experience of stuttering generates emotions such as frustration, fear, and anger in everyone who stutters. In other words, emotional arousal may cause stuttering, but stuttering may also cause emotional arousal.
I will begin the discussion of emotion and stuttering with a review of the link between stuttering and anxiety, and then I will describe the evidence suggesting that many people who stutter have more emotionally sensitive temperaments than fluent speakers.
Anxiety and Autonomic Arousal
The average listener may think that people stutter because they are nervous. Scientists, following up on this impression, have used such terms as “anxiety,” “autonomic arousal,” and “negative emotion” to specify the emotional states that may cause or accompany stuttering. The broad term “anxiety” generally describes a state of alert concern about a future event. When the term “autonomic arousal” is used in a similar way, it denotes activation of the sympathetic nervous system, which prepares the body for action, such as in the fight-or-flight response. Research about anxiety and stuttering has been going on for more than 50 years. Recently, Bloodstein and Ratner (2008) reviewed more than a dozen studies comparing stutterers and nonstutterers on various measures of anxiety and found that more than half of the studies revealed no differences. However, a substantial number of them did find stutterers to be more anxious on some measures. It is possible that there is a subgroup of stutterers who are more anxious, and studies that happen to have more individuals from this subgroup in their study find the stuttering group to be more anxious. No doubt some of the studies of social anxiety may reflect the fact that years of stuttering can cause a person who stutters to be more anxious in social situations.
Several studies related to anxiety and stuttering are described in detail in Chapter 3. In summary, research has shown that high levels of anxiety produced by the threat of electric shock can produce stuttering-like behaviors in typical speakers. Moreover, there is evidence that changes in speech-related physiology occur in stutterers but not nonstutterers under conditions of anxiety. Finally, several studies using physiological measures of anxiety showed stutterers and nonstutterers to be equally anxious, but only stutterers showed the effects in their speech in terms of increased disfluencies.
Many of us who work with children who stutter have often heard parents describe their children as particularly sensitive. Upon questioning, these parents frequently say that even before stuttering began, the child was more easily upset by changes in routine or was shyer with strangers than her siblings. These emotional and behavioral characteristics may be a part of the child’s inherited physiology or temperament. Some children seem to be born with sensitive or inhibited temperaments and are more likely to react to new people and novel situations with increased muscle tension and physiological signs of stress (Kagan, Reznick, & Snidman, 1987).
An important early conceptualization of stuttering, temperament, and anxiety was presented in Brutten and Shoemaker’s (1967) classic book The Modification of Stuttering. Rather than using the term “temperament,” they referred to “individual differences in conditionability and autonomic reactivity.” They suggested that some individuals have predispositions to stutter because they are constitutionally more likely to have an anxiety-based speech breakdown under stressful conditions. Moreover, these individuals are also thought to be more conditionable, making it more likely that initial breakdowns under stress will escalate into highly learned stuttering behaviors.
Following Brutten and Shoemaker, a number of authors have speculated about the possible importance of considering this kind of reactive temperament in gaining a better understanding of the nature of stuttering (e.g., Bloodstein, 1987, 1995; Bloodstein & Ratner, 2008; Conture, 1991; Guitar, 1997, 1998, 2000; Peters & Guitar, 1991; Walden, Frankel, Buhr, Johnston, Conture, & Karrass, 2012). A reactive temperament, for example, might trigger increased physical tension in a child when she is disfluent and thus create a learned cycle of disfluency begetting more severe disfluencies, leading to chronic stuttering. On the other hand, a placid temperament in an equally disfluent child might allow her to stay relaxed, ignore the disfluencies, and thereby outgrow early stuttering.
Data on the sensitivity of people who stutter are meager, but what is available is described in Chapter 3. Questionnaire studies have found indications that both adults and children who stutter are more sensitive than nonstutterers. This has been corroborated by at least one study of a physiological measure of sensitivity.
With evidence in hand that at least some individuals who stutter have more sensitive temperaments, we need to ask how this may shed light on the disruption of fluency by emotion. Psychologists who study temperament have looked carefully at the regulation and expression of emotion in persons with sensitive temperaments. There is good evidence from studies of both normal and brain-damaged patients that the regulation of emotion is a lateralized function (Kinsbourne, 1989; Kinsbourne & Bemporad, 1984). Emotions regulated by the left hemisphere appear to motivate such behaviors as approach, exploration, and action, whereas emotions regulated by the right hemisphere motivate behaviors such as avoidance, withdrawal, and the arrest of action. Studies of electrical activity in the brain indicate that individuals with sensitive temperaments are right hemisphere-dominant for emotionally based behaviors (Ahern & Schwartz, 1985; Calkins & Fox, 1994). This means that if stutterers are temperamentally reactive as a group, they may have an inborn proclivity toward behaviors motivated by righthemisphere emotions–avoidance, withdrawal, and the arrest of action.
How this may affect speech is not yet clear, but Webster (1993b) speculates that when individuals who stutter are emotionally aroused, then right-hemisphere proclivities, such as avoidance and withdrawal, could affect their left-hemisphere SMAs, interfering with planning and initiation of speech. My own speculation about the relationship between emotion and stuttering is that one important aspect of right hemisphere–dominant, emotionally based behaviors is the arrest of ongoing behavior. This phenomenon is especially well described by Jeffrey Gray (1987), a psychologist who has studied the central nervous system’s response to stress. He proposes that when an individual experiences fear or frustration, a behavioral inhibition system in the brain increases three distinct forms of behavior: (a) freezing, which involves widespread muscular contractions that produce tense and silent immobility; (b) flight; or (c) avoidance. It is possible that such behaviors may be manifested in speech by both core behaviors (repetitions, prolongations, and blocks) and secondary behaviors (escape and avoidance).
Indirect support for this possibility may be findings by Kagan, Reznick, and Snidman (1987) that more sensitive children manifest their reactivity by generating higher levels of physical tension, particularly in laryngeal muscles, when they are speaking in unfamiliar or threatening situations. I suspect that some children who are both sensitive and have vulnerable motor speech systems may respond to early repetitive disfluencies with increased tension, especially in the laryngeal region. This tightening may further interfere with speech, producing the abruptly terminated repetitions, prolongations, and blocks that develop in many children when stuttering persists. Other children who are highly sensitive and predisposed to have motor speech breakdowns may begin their disfluencies with tense prolongations and blocks in response to emotionally difficult situations. The heterogeneity of individuals who stutter and their unique patterns of stuttering will be discussed further in upcoming chapter sections on developmental factors and learning.
Brain imaging studies have shown extensive activity during stuttering in an area called the right insula, shown in Figure 2.3 (Fox, 2003), and the anterior cingulate cortex (Braun, Varga, & Stager, 1997; Braun, Varga, Stager, Schulz, & Selbie et al., 1997; De Nil, Kroll, Kapur, & Houle, 2000). Both of these areas have strong connections with the amygdala (Allman, Hakeem, Erwin, Nimchinsky, & Hof, 2001; Habib, Daquin, Milandre, Royere, & Rey et al., 1995), which is a major structure in fear conditioning (LeDoux, 2002). It is possible that some of the right-hemisphere activity heightened during stuttering and reduced during induced fluency may reflect negative emotional arousal. My reasoning is that, first, many of the studies reviewed for this chapter suggest that stutterers’ speech planning and production are localized in right-hemisphere regions homologous to Broca’s, Wernicke’s, and interconnecting areas. Second, emotions lateralized to the right hemisphere in the human brain are those associated with fear–avoidance, escape, and arrest of ongoing behavior (Kinsbourne, 1989). Third, because strong emotions tend to dominate the neural processes in surrounding areas (LeDoux, 2002), these emotions may disrupt ongoing speech processing in ways analogous to how they affect all behavior, including avoidance behaviors, escape behaviors, and blocks.
The section you have just read–on emotions and stuttering–suggests that emotions play a major role in the development of stuttering. In some cases, they may also play a role in the cause. All of this information needs to be pulled together into a unified description that explains more precisely how emotion influences stuttering (and vice versa), as well as how individuals may differ in this respect. Some aspects of a unified view of stuttering and emotions are suggested in a model proposed by Walden and colleagues (2012). Although their view (“dual diathesis-stressor model of stuttering”) incorporates both emotional variables and speech-language variables, I will only describe what they hypothesize about emotions. They suggest that children who stutter may have constitutional predispositions (diatheses) that make them highly emotionally reactive to novel stimuli. This predisposition will be greater or lesser in different children. For the predisposition to be “activated,” the child must encounter some environmental stress. Thus, the child may stutter more or less in any given situation, depending on the stress he experiences and the degree of predisposition he has. The stimulus the child is reacting to in this case is the experience of having some difficulty speaking (disfluencies or other speech disturbances). The child’s emotional reaction to the difficulty will increase the difficulty in a cyclical fashion. Of course, any given child will have other predispositions, such as language or speech deficits that will interact with the emotional diathesis.
It is hoped that future research will provide evidence to support these hypotheses and other theoretical perspectives on stuttering. New theoretical frameworks need to include descriptions of how emotion may affect recovery from stuttering–via treatment or natural recovery–and suggest whether recovery can be improved by treating emotions as well as stuttering behaviors. One example of new research that will help build models of the connection between stuttering and emotion is the research of Kang et al. (2010) on genes associated with stuttering, described earlier in this chapter. This work has identified one gene (GNPTG) that is associated with both motor control and emotional regulation. Exploration of how individuals with this gene respond to treatment and how treatment may be adjusted to modify the effect of this gene would be a vital step forward.
Complete theoretical models of stuttering must, of course, incorporate all of the constitutional factors described in this chapter–not only genetics and emotion (Table 2.7) but also brain structure and function, sensory processing, sensory-motor control, and language. In addition, developmental and environmental factors must be included as well.
Table 2.7 Summary of Emotional Factors in Stuttering

If you are talking with a person who stutters or a parent of a child who stutters, you may want to summarize current thinking about the causes of stuttering. I think it would be safe to say something like this:
“Recent scientific findings about stuttering suggest that individuals who stutter have a slightly different brain structure and function than fluent speakers. Because of heredity or possibly early injury, the brains of people who stutter seem to be a little less efficient at bringing together all the elements of spoken language–like what words to use and how to put those together into a sentence–at just the right time and speedily enough for rapid, fluent speech. This doesn’t happen all the time, of course. But when a person is saying something complicated or is in a hurry or is really excited or upset, then being fluent is a little harder.”
In this chapter, I have reviewed the evidence for a constitutional basis for stuttering, and in the next chapter, I’ve given more detail on the research findings that make up this evidence. In Chapter 4, I will discuss developmental and environmental factors, which may interact with the constitutional factors to produce the disorder. In the subsequent chapter, I will integrate these findings to provide a comprehensive view of how all of the factors may combine to precipitate stuttering in an individual.

• Stuttering appears to have a genetic basis in many individuals. However, twin studies and adoption studies confirm that genes must interact with environmental factors for stuttering to appear.
• Recent research identifies some genes associated with stuttering in some individuals.
• Stuttering may have its etiology in congenital factors for some stutterers. These mayinclude physical trauma at birth or in utero, cerebral palsy, retardation, and emotionally stressful situations.
• Slightly more boys begin to stutter than girls, but girls are more likely to recover, soby school age and beyond, there are many more boys who stutter than girls.
• Early childhood stuttering may be either transitory, in which the child recovers naturally within 18 months, with no or minimal treatment, or persistent, in which the child, if not treated, stutters three years or more.
• Persistent and transitory stuttering appear to be the result of a common geneticfactor (either a single gene or several), but the persistent form of stuttering probably has additional genetic factors that impede recovery.
• Natural recovery from stuttering appears to be associated with the following factors: (a) good scores on tests of phonology, language, and nonverbal skills; (b) either no family history of stuttering or family members who had natural recovery from stuttering; (c) early age of onset of stuttering; and (d) being a girl.
• Brain imaging studies of adults who stutter have shown various anomalies duringspeaking and especially during stuttering. One anomaly is overactivation in rightbrain areas homologous to left-hemisphere speech and language structures typically used by nonstutterers. Another anomaly is deactivation in the left auditory cortex.
• Neuroanatomical differences seen via brain imaging include (a) anomalies in theplanum temporale (related to auditory processing) and in gyri (raised areas on brain’s surface) in speech and language areas and (b) less dense fiber tracts connecting speech perception, planning, and execution areas.
• Inducement of short-term or long-term fluency in stutterers is accompanied by decreases in right-hemisphere activations and increases in activation of lefthemisphere speech, language, and auditory areas.
• On tasks of sensory processing, stutterers have less accurate and slower processing,particularly of auditory stimuli, and lack of left-hemisphere dominance for processing.
• Greatest performance deficits occur when linguistically complex stimuli are used.
• Masking and other changes in auditory feedback create temporary fluency, suggesting that distortions, deficits, or delays in auditory feedback may be associated with stuttering.
• On tasks of sensory-motor control, stutterers demonstrate slower reaction times,especially when stimuli are linguistically more complex. Stutterers are slower, less accurate, and less left hemisphere-dominant when performing sequential motor tasks and auditory-motor tasks.
• When there is a greater linguistic load, stutterers’ speech motor systems are morevariable; greater linguistic load is also associated with more stuttering.
• Stutterers do not appear to be more anxious than nonstutterers, but there is evidencethat when their autonomic arousal levels are high, more stuttering is likely to occur.
• As a group, stuttering children and adults appear to have a more sensitive temperament. This sensitivity may be associated with more physical tension in speech musculature for some individuals.

1. How does each of the areas–family studies, twin studies, and adoption studies–provide evidence that stuttering is inherited?
2. A couple comes to you for advice. They tell you they are thinking of havingchildren but are worried because each has a relative who stutters. What more information would you like to get from them? What would you tell them about the likelihood that they would have a child who stutters and whether they should be concerned?
3. How do studies provide evidence that stuttering is a product of both heredity andenvironment?
4. How would you summarize the brain imaging studies to someone who is not aprofessional in our field?
5. Why do almost all the brain imaging studies use right-handed males as participants?
6. Researchers have found many differences between groups of stutterers and nonstutterers. Why can’t we say that these differences cause stuttering?
7. Explain how the deficits in sensory processing in people who stutter could berelated to the actual behaviors of stuttering?
8. What are the differences between sensory processing and sensory-motor control?
9. Do you think difficulty with language processing may be a cause of stuttering forsome individuals? Why or why not? How might language deficits be related to stuttering?
10. Describe the relationships between emotion and stuttering.
11. What research finding in this chapter do you think has the most relevance for thetreatment of stuttering? Defend your answer.
1. Talk to someone who stutters and plot out his or her family tree, noting relatives who stutter and relatives who have other speech, language, or learning problems.
2. Make a family tree of your own relatives indicating which, if any, currentlyhave or have had speech, language, hearing, or learning disabilities. Describe how you got the information and what the disabilities are.
3. On which side of the brain do you process speech and language? Find out howyou could ascertain this information by asking speech-language pathology
researchers or audiologists you know if they have tests you could take to find out. If this doesn’t lead to a test for this kind of laterality, search the internet for self-administered tests, which tell you whether you are more “left-brained” or more “right-brained.”
4. Use a digital stopwatch (how fast can you turn it on and off?) or similarinstrument to determine your reaction time. Try this under many different conditions, such as at several times during the day and when sick or tired versus when feeling alert. Determine what variables affect your reaction times, and see if it is true for other people. Using this information, suggest why different studies of reaction times in stutterers get different outcomes.
5. Find a temperament test, such as those available online, and take it yourself.
Do you think the results accurately describe you?
Bloodstein, O., & Ratner, N. (2008). A handbook on stuttering (6th ed.).
Clifton Park, NY: Delmar Learning.
This is the most recent edition of a classic reference book on stuttering. It provides a thorough update of “the most important research in stuttering.” Moreover, it is quite easy to read.
Maassen, B., Kent, R., Peters, H., van Lieshout, P., & Hulstijn, W. (Eds.) (2004). Speech motor control in normal and disordered speech. Oxford:
Oxford University Press.
This book contains updated chapters by many scientists who presented their findings at a conference on speech motor control in 2001. Although not all of the contents are directly related to stuttering, there is much of great interest to clinicians and researchers interested in the neurophysiological bases of speech and stuttering.
Neumann, K., & Euler, H. (2010). Neuroimaging and stuttering. In B. Guitar & R. McCauley (Eds.), Stuttering Treatment: Established and Emerging Approaches (pp. 355–377). Baltimore: Lippincott Williams & Wilkins.
This chapter begins with a history of brain imaging and stuttering and then describes the most important findings in structural and functional brain imaging related to stuttering. This is followed by a section on neuroimaging findings before and after treatment, a specialty of the authors.
Van Riper, C. (1982). The nature of stuttering (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.
Although somewhat outdated, this book reviews an impressive amount of world literature on stuttering, from as long ago as the 20th century B.C. In a synthesis of the research, Van Riper presents his venerable hypothesis that stuttering is a disorder of timing.
Yairi, E. & Ambrose, N. G. (2005). Early childhood stuttering. Austin, TX:

The authors give an in-depth description of the results of 14 years of research on the development of stuttering conducted at the University of Illinois. Chapters are devoted to the onset and development of stuttering, characteristics of children’s disfluency, genetics, and cognitive, psychosocial, and motor factors in stuttering. Elaine Paden and Ruth Watkins contributed chapters on phonological and language abilities of children who stutter, respectively. Like Wendell Johnson’s magnum opus The Onset of Stuttering, this book reflects a monumental effort focused on childhood stuttering.

Research Findings about
Constitutional Factors in Stuttering

Hereditary Factors
Family Studies
Twin Studies
Adoption Studies
Congenital and Early Childhood Trauma Studies
Brain Structure and Function
Brain Structure Differences in People Who Stutter
Brain Function Differences in People Who Stutter
Electroencephalographic Studies
Cerebral Blood Flow Studies
Positron Emission Tomography Studies and Beyond
Brain Overactivation During Stuttering
Overactivation of Right-Hemisphere Cortical Areas During Stuttering Overactivation in Midbrain Areas
Underactive Brain Areas in Stuttering
Underactivity in Speech Motor Areas
Underactivity in Auditory Areas
Sensory and Sensory-Motor Studies
Brain Electrical Potentials Reflecting Auditory Processing
Dichotic Listening Tests
Auditory, Tactile, and Proprioceptive Feedback
Sensory-Motor Control
Fluent Speech
Nonspeech Motor Control
Language Factors
Emotional Factors
Anxiety and Autonomic Arousal
After studying this chapter, readers should be familiar with details of the research literature relevant to constitutional factors in stuttering.
In Chapter 2, I sketched the “big picture” of constitutional factors, summarizing the major themes so that you could see how an individual’s neurological makeup could explain many of the behaviors of stuttering. In this chapter, I’ll fill in the details from the multitude of scientific studies that have been carried out on stuttering for almost 100 years.
Family Studies
The first “modern” reports on the genetics of stuttering were published by a group of researchers in Newcastle, England (Andrews & Harris, 1964; Kay, 1964). They investigated the family histories of 80 stuttering children and compared them with the families of nonstuttering children. These researchers found that (a) children who stuttered had far more stuttering relatives than did children who didn’t stutter; (b) male children were at higher risk for developing stuttering than female children; and (c) female children who stuttered were more likely to have stuttering relatives than were male children who stuttered. Thus, a pattern of family occurrence emerged, providing evidence for a possible genetic explanation of stuttering. These results supported a model in which stuttering was transmitted by either a single gene or a combination of several genes contributing different factors. This early insight into the possibility of multiple genes is supported by more recent work. My own working hypothesis is that chronic or persistent stuttering is the result of genes affecting not only speech motor control but also language and temperament.
Ten years after the Newcastle studies, researchers at Yale University conducted further family studies of stuttering (Kidd, 1977; Kidd, Kidd, & Records, 1978; Kidd, Reich, & Kessler, 1973). Using the data from England (Kay, 1964) combined with new data they gathered themselves in the United States, the Yale researchers were able to develop statistical models, which predicted patterns of inheritance. They found, as in Kay’s earlier study, that males were more likely to stutter than females and that females who stuttered were more likely to have relatives who stuttered. Kidd (1984) concluded that these patterns were best explained by an interaction between the environment and a combination of several genes.
The Newcastle and Yale family studies focused on children and adults, most of whom had been stuttering for several years. Researchers at the University of Illinois used a different approach. Ambrose, Yairi, and Cox (1993) studied the family histories of 69 very young children who had just been diagnosed with stuttering, including individuals who would recover from stuttering and those who would persist. The Illinois group found that two-thirds of these children had relatives who stuttered and that, as in earlier studies, more male relatives than female relatives stuttered. Unlike past studies, however, these researchers found that male and female children who stuttered had similar chances of having relatives who stuttered. This difference is likely to have come from the fact that females in the Kay and Kidd studies were older, with persistent rather than transient cases of the disorder. Kay and Kidd may have found that their females had more relatives who stuttered because they studied only those females whose stuttering persisted, and for a female (who is less likely to stutter than a male) to persist in stuttering, she can be assumed to have inherited more genetic material. Thus, she would naturally have more relatives who stuttered; her family would have a heavier “genetic loading.” On the other hand, many of Ambrose, Yairi, and Cox (1993) very young female subjects recovered quickly from stuttering and thus may have had lower genetic loadings and fewer relatives who stuttered.
Kay’s (1964) and Kidd’s (1984) hypotheses that stuttering may be transmitted by several genes rather than a single gene received support from another study by the Illinois group that investigated differences between those young children who recover from stuttering without treatment and those who persist. Ambrose, Cox, and Yairi (1997) analyzed the family trees of 66 children who were identified soon after the onset of stuttering. The children were followed for several years and eventually grouped into those who persisted in stuttering and those who recovered. The researchers found that the sex ratios of the two groups were quite different. The male: female ratio was 7:1 in the persistent group but about 2:1 in the recovered group, indicating a much higher percentage of boys in the persistent group. This provides more evidence that girls are more likely to recover than boys.
A second finding was that persistence tended to run in families. In other words, children who did not outgrow their stuttering were likely to come from families in which relatives who stuttered also persisted in their stuttering. Conversely, children who recovered were likely to come from families in which relatives who initially stuttered became fluent when they grew older. Further analysis of their data led the authors to propose that persistent and recovered stuttering are transmitted by the same major gene or genes, but that those individuals whose stuttering persists have additional genetic factors that hamper recovery. It is also possible that those who recovered have additional genetic factors that facilitate recovery. A contrasting view is given by Viswanath, Lee, and Chakraborty (2004), whose studies of persistent stutterers led them to hypothesize that persistent and recovered stuttering are two genetically different disorders.
The researchers in Illinois examined a number of genetic and nongenetic factors that might predict recovery or persistence and thereby might be useful in deciding which children are in immediate need of treatment. In an early study, Yairi, Ambrose, Paden, and Throneburg (1996) found that predictors of recovery include (a) good scores on tests of phonology, language, and nonverbal skills; (b) family members who had recovered from stuttering; and (c) early age of onset of stuttering. Some of the factors that impede recovery, such as problems in phonology or language, might be determined by other genes that may accompany a gene that is related to the initial onset of stuttering. In later studies (e.g., Yairi & Ambrose, 1999), the Illinois group refined this list of factors; this refined list, which includes eight factors, was presented in Chapter 1 in the section titled “Recovery without Treatment.”
Before we leave the topic of family studies, it is worth mentioning several criticisms that have been made about genetic studies in reviews by Felsenfeld (1997) and Yairi, Ambrose, and Cox (1996). Many of the genetic studies used no matched control group, but instead relied on incidence figures from other studies. Other problems in past studies include the inadequacy of a definition of stuttering when searching for stuttering among relatives of children or adults who stutter and relying on testimony of others, rather than direct assessment to determine whether or not a relative stutters. These researchers suggest that future studies (a) look for subgroups of stutterers that may have different genetic etiologies; (b) examine family members who don’t stutter to find factors that may resist stuttering; and (c) search for environmental factors that may interact with genetic factors to precipitate or maintain stuttering.
Twin Studies
Twin studies of stuttering have shown that the disorder occurs much more often in both members of identical twin pairs than in both members of fraternal, same-sex twin pairs (Andrews, Morris-Yates, Howie, & Martin, 1991; Felsenfeld, Kirk, Zhu, Statham, Neale, & Martin, 2000; Howie, 1981; Luchsinger, 1944; Seeman, 1937). To use the vocabulary of genetics, there is higher “concordance” of stuttering in identical than in fraternal twins. This supports the hypothesis that stuttering is inherited, but it doesn’t reveal what exactly is inherited. How does a gene (or several genes) affect a child’s speech so that stuttering results? No one is sure.
In addition to providing evidence of genetic factors in stuttering, twin studies demonstrate that heredity does not work alone. In one of the twin studies previously cited, although there was higher concordance for stuttering among identical twins, some pairs were discordant (Howie, 1981). Specifically, Howie found that in six of the 16 identical twin pairs, one twin stuttered, but the other didn’t. This means that even though both members of the twin pair had the same genetic inheritance, only one of them stuttered, indicating that genes alone do not explain stuttering. However, this may not be surprising because genes must interact with the environment to produce their effects (LeDoux, 2002). A gene might not express itself in stuttering unless, for example, there is some kind of prenatal or postnatal stress on the child. In the case of stuttering, where there appears to be several genes working together to produce a chronic disorder, the situation is even more complex because several genetic tendencies may need to interact with different aspects of the child’s internal and external environment to create stuttering. No wonder there were six discordant pairs in the Howie study!
An estimate of the relative proportions of genetic and environmental influences was suggested in a later study involving 3,810 unselected twin pairs (Andrews et al., 1991). They were deemed “unselected” because they were all part of the Australian
Twin Registry rather than a population selected because of stuttering, thus making the sample less likely to be influenced by “ascertainment bias.” This kind of bias might occur, for example, if subjects were found through newspaper ads; only individuals who read the ads in newspapers and were motivated to be in the study would participate in the study. Analyses of stuttering in these 3,810 unselected twin pairs estimated that 71 percent of the variance (the probability of whether or not one would stutter) is accounted for by genetic factors, and 29 percent is accounted for by the individual’s environment (including factors influencing the fetus, as well as factors after birth). Felsenfeld and colleagues (2000) recently followed this study up by contacting a group of twins in the Australian Twin Registry different from those studied by Andrews and colleagues (1991). Felsenfeld and colleagues (2000) screened 1,567 pairs and 634 individuals, using questionnaires and telephone interviews. They found 17 monozygotic and eight dizygotic twin pairs who were concordant for stuttering (both twins stuttered at some time in their lives) and 21 monozygotic and 45 dizygotic twins who were discordant for stuttering (only one of the twin pairs ever stuttered). Statistical analyses estimated that “additive genetic effects” (the effects of different genes working together) accounted for 70 percent of the variance and that an individual’s unique environment (influences on one of the twins but not the other, such as illness) accounted for 30 percent of the variance. These proportions are essentially the same as those found by Andrews and colleagues (1991) and support current thinking that genes and environment interact to set the stage for stuttering.
In a study of 1,896 twin pairs in Japan, Ooki (2005) compared concordance in identical and fraternal twins. Using sophisticated statistical techniques, Ooki determined that the proportion of genetic influence on stuttering in males was 80 percent in males and 85 percent in females. In other words, for both males and females, genetic factors were strongly implicated–even more so than in the Andrews and team (1991) and Felsenfeld and team studies (2000). It is interesting that the females showed slightly more genetic influence than the males. Perhaps this echoes the evidence that females have some resistance to stuttering. For a female to develop stuttering, more genetic influence or “genetic loading” is needed.
A study by Dworzynski, Remington, Rijsdijk, Howell, and Plomin (2007) discovered interesting differences between a group of twins who recovered from stuttering (n = 950) and a group who persisted (n = 150). In the recovered group, concordance for stuttering was 40 percent for identical twins and 20 percent for fraternal twins. However, in the persistent stuttering group of twins, the concordance was 19 percent for identical twins and 0 percent for fraternal twins. This suggests that the genetics of persistent stuttering are complex. As we noted previously, the family studies of Ambrose, Cox, and Yairi (1997) indicated the possibility that while recovered and persistent stuttering are transmitted by the same major gene(s), persistent stuttering itself may have additional genetic factors that make recovery more difficult. The findings of Dworzynski and colleagues (2007) may support this supposition because there is so little concordance in the persistent fraternal twin group –meaning that to get concordance, the same array of multiple genes must be transmitted, far less likely in fraternal twins.
Van Beijsterveldt, Felsenfeld, and Boomsma (2010) conducted a study using a very large participant pool: 105,000 twin pairs at age 5. They bypassed the usual problem of parent identification of their children as stuttering by asking parents merely to estimate the frequency of repetitions, prolongations, and blocks they observed in their children’s speech. Children were categorized by the experimenters as “probably stuttering” or “high nonfluency,” or as having typical speech. Concordance for probable stuttering was higher in identical twins, supporting the genetic/heritability hypotheses. It was notable that high nonfluency also appeared to be genetically based.
In an interesting aside, Bloodstein and Ratner (2008) questioned the assumption that influence on stuttering that was not accounted for by genetic factors must be attributed to environmental factors. Research using animal models suggests that identical twins sometimes have discordance for certain traits not because of environmental influences but because of variations in the way two identical embryos develop. Some of these variations may be related to “epigenetics,” non-DNA factors that are inherited and influence the expression of genes into specific behaviors or phenotypes. Two examples in humans are Angelman syndrome and Prader-Willi syndrome. These two syndromes appear very different but are the result of deletion of the same DNA sequence (gene) in the chromosome. The difference between the two syndromes is simply whether the chromosome is inherited from the mother or the father. In the two syndromes, the chromosome inherited from the mother has a different “tag” or expression controller than the chromosome inherited from the father. We describe this difference as an epigenetic factor because the genes are identical, but how they are triggered or expressed is different (Kempf & Weinberger, 2009). Readers wishing to understand recent developments in genetics, epigenetics, and other principles of gene variability may want to view “Ghost in Your Genes” a public broadcasting system NOVA program on genetics located at
Adoption Studies
Because the birth records of adopted children are difficult to obtain, studies of adopted stutterers are rare. Bloodstein (1961b) and Bloodstein and Ratner (2008) presented information obtained from 13 adopted stutterers whom Bloodstein interviewed about stuttering in their adoptive families (information on their biological families was not available). Four of the 13 stutterers reported having relatives who stuttered in their adoptive families, which is higher than would be expected by chance. This small sample, without data from biological families, supports the possibility that environmental factors may have an effect. If the relatives in the adoptive family were key figures, such as a parent or older sibling who was close to the child, this would be stronger evidence for the influence of the environment on stuttering. Unfortunately, this information is not available.
Felsenfeld (1997) reported some preliminary data on a small sample of adopted children who had speech disorders (primarily stuttering) and for whom data were available from both adoptive and biological families. These data indicated that a history of stuttering in the biological families was slightly more predictive of disorders in these children than was stuttering in the adoptive family.
Again, the evidence suggests that both genetic and environmental factors influence whether or not a child will stutter and that genetic inheritance appears to contribute more strongly.
A number of researchers are currently looking for genes that may predispose children to stuttering. Dennis Drayna (1997) at the National Institutes of Health has begun genetic linkage studies using large numbers of families to try to isolate genes for stuttering. Linkage analysis compares the chromosomes of family members who have a trait (the appearance of a trait in a person is called a “phenotype”) with those of family members who do not. In this way, the chromosomal location of the stuttering gene or genes can be identified. Drayna has been studying families in which there is more than one individual who stutters. His work has brought him into contact with a family in Cameroon, Africa, in which 42 of 100 family members stutter. Preliminary results suggest genes on chromosome 18 may be related to stuttering, a set of genes that control intercellular communication (Shugart, Mundorff, Kilshaw, Doheny, & Doan et al., 2004).
Cox and Yairi (2000) have been studying individuals who stutter in a North Dakota community of Hutterites. This group is of interest because they do not marry outside the community, resulting in a homogeneous gene pool. These researchers have identified three chromosomes (numbers 1, 13, and 16), which may include the genes involved in stuttering. Riaz, Steinberg, Ahmad, Pluzhnikov, and Riazuddin et al. (2005) added another chromosome to the list of those that may contain a gene associated with stuttering. The researchers studied 46 Pakistani families who were highly inbred, increasing the probability that if a particular chromosome appears to be associated with stuttering, many stuttering individuals in different families would also have the same chromosome because they are all so highly interrelated. Their results suggest that chromosome 12 in these families is strongly linked to stuttering, and chromosome 1 is a little less strongly linked. The authors point out that their results and those of other groups suggest that all stuttering is not likely to be the result of the same chromosome location; there may be several different locations that can contribute to stuttering.
Not long after the Riaz team (2005) study, Suresh, Ambrose, Roe, Pluzhnikov, &
Wittke-Thompson et al., (2006) published a genome linkage study of 100 families of European descent in the United States, Sweden, and Israel. Their research added some new ways of looking at genetic associations with stuttering. First, they compared chromosome locations associated with stuttering for individuals who had stuttered sometime in their lifetimes (some of whom had recovered, some of whom were still stuttering) versus a subgroup of these individuals who persisted in stuttering. Chromosome 9 appeared to be the location of a gene related to stuttering at sometime in one’s lifetime; chromosome 15 was linked only to those whose stuttering was persistent. This is support for the idea that whether an individual naturally recovers from stuttering or not is determined, at least in part, by genetic makeup. It also supports the finding that one predictor of natural recovery is family history of recovery from stuttering (Yairi & Ambrose, 2005).
A second new twist in this study was comparing chromosome locations for stuttering that were specific to males compared to females. The researchers found that in males, stuttering was linked to a location on chromosome 7, and in females, it was linked to chromosome 15. They noted that chromosome 7 has also been associated with specific language impairment and autism. A secondary analysis of those families whose stuttering was associated with chromosome 7 found strong evidence of a linkage of stuttering to chromosome 12. This is of interest because of the evidence for chromosome 12 in the findings of Riaz and colleagues (2005), who examined a group of families in Pakistan who were culturally and geographically very different from those in the Suresh team (2006) study.
Recently, a follow-up study was conducted with those families in Pakistan that showed a link between stuttering and chromosome 12. Kang, Riazuddin, Mundorff, Krasnewich, Friedman, and colleagues (2010) studied the 123 Pakistanis who stuttered as well as 270 individuals in the United States and England who stuttered. They also used a control group of 372 individuals in these three countries who didn’t stutter. Mutations of three genes (the genes were called GNPTAB, GNPTG, and NAGPA) on chromosome 12 were found to be associated with stuttering. Some individuals showed mutations of gene GNPTAB, others had mutations on GNPTG, and still others had mutations on NAGPA. The work of all three genes is related to controlling enzymes in a cell’s lysosome structure, the part of a cell involved in recycling cell waste products. It is important to note that there are known genetic disorders (e.g., mucolipidoses) of this waste recycling process that affect joint, skeletal, and other body components. These disorders affect brain development, resulting in delays in movement coordination, and some forms of mucolipidoses are accompanied by speech problems (National Institute of Neurological Disorders and Stroke, 2011). Additionally important is the fact that one of the genes, GNPTG, is associated with motor control and emotional regulation by way of the gene’s expression in the hippocampus and in the cerebellum.
One of the first studies to look closely at stutterers who had no family history of stuttering, West, Nelson, and Berry (1939), examined a sample of 204 people who stuttered and found that 100 of them reported no family history of stuttering. Of these 100, 85 reported congenital factors that may have been related to the onset of stuttering. These factors included infectious diseases, diseases of the nervous system, and injuries–all reported to have occurred just prior to stuttering onset, although the exact proximity to onset was not reported. Thus, these congenital factors may have created a predisposition to develop stuttering.
A later study by Poulos and Webster (1991) found that 57 of the clients in a clinic sample of 169 adults and adolescents who stuttered reported no family history of stuttering. Of these without family histories, 37 percent reported congenital factors that may have been associated with the onset of stuttering, whereas only 2.4 percent of the clients having a positive family history of stuttering reported such factors. The factors reported included anoxia at birth, premature birth, childhood surgery, head injury, mild cerebral palsy, mild retardation, and experiencing intense fear.
A third study that began with a larger group of stutterers and then looked for possible etiological differences was reported by Alm and Risberg in 2007. At first they divided the group in half, using a scale that measured tendencies toward attention deficit hyperactivity disorder (ADHD). Looking closer at the two subgroups, they discovered that those with high ADHD tendencies had more neurological lesions prior to stuttering onset and also had less family history of stuttering. This again is support for the hypothesis of two different predispositions contributing to stuttering–genetic inheritance and brain injury.
There were two studies that looked at a brain-injured population and assessed whether there were more individuals who stuttered than in the general population. In the first, Böhme (1968) examined a sample of 313 individuals who had sustained brain damage at birth or in early childhood; 24 percent of those 313 developed stuttering (compared to 5 percent in the general population). Unfortunately, no information is given about family history of stuttering in those who developed it, but the implication is that congenital or early childhood brain injury can often result in stuttering. In a similar study, Segalowitz and Brown (1991) surveyed more than 600 high school students to ascertain how many had experienced head injury during their childhoods. Of the students, 92 reported head injury, and of those, nine (about 10 percent) reported having been diagnosed with stuttering. However, it was not clear that the stuttering appeared after the head injury. The authors found that there was a significant relationship between having a head injury and being diagnosed with stuttering, particularly for those children who were unconscious for a period of time after the head injury. Again, there is no information as to whether some of the children who stuttered and had head injuries also had family histories of stuttering.
The fact that neurological or psychological traumas may be associated with childhood stuttering in those without family histories of stuttering is not surprising. Adult onset of stuttering is often associated with head injury, neurological disease, stress, or psychological trauma as I will describe in Chapter 15. Thus, mechanisms similar to those precipitating adult onset may be involved in childhood stuttering in the absence of family history of stuttering, but this possibility raises as many questions as answers. Why would some children (and adults), but not others, begin to stutter as a result of intense fear or brain injury? Which brain structures and functions affected by head injury and neurological disease result in stuttering? How are they similar to and how do they differ from the effects of inheriting a predisposition to stutter?
In respect to the last question, a number of investigators have looked at whether stutterers who have family histories of stuttering showed any differences in their stuttering behavior, such as severity, compared to those without family histories. Andrews and Harris (1964) and Kidd, Heimbuch, Records, Oehlert, and Webster (1980) found no differences in the stuttering of those with and without family histories of stuttering. However, Janssen, Kraaimaat, and Brutten (1990) looked at a wider variety of speech and language-related variables in several different age groups. They found several significant differences (as well as similarities) between the group with family histories of stuttering and the group without such histories. In terms of similarities, the groups were not significantly different in responsiveness to treatment, reading ability, or speech-related anxiety. However, when stuttering behaviors were examined closely, those with family histories of stuttering showed more prolongations and blocks (silent prolongations) than those with no history of stuttering, although the frequency of repetitions was the same for both groups. Another difference between the groups was that those with positive family histories for stuttering showed significantly longer durations of voiced segments of speech and significantly greater variability in length of unvoiced segments during fluent speech than those with no family history of stuttering. The authors summarized this finding by suggesting that the stutterers with positive family histories were slower and more variable in their fluent speech.
It could be concluded from these studies that individuals with family histories of stuttering have inherited greater neuromotor instability than those without family histories of stuttering–an instability that produces more prolongations and blocks and that may require the individual to speak more slowly to maintain fluency. This is not to say that those without family history of stuttering did not inherit the predisposition to stutter. Their family histories may contain other speech-related deficits (e.g., articulation problems), and their underlying neuromotor anomalies may result in stuttering, but it is stuttering that is more characterized by repetitions than prolongations or blocks.
Brain Structure Differences in People Who Stutter
In an early study of brain anatomy, researchers at Tulane University in New Orleans (Foundas, Bollich, Corey, Hurley, & Heilman, 2001) used magnetic resonance imaging (MRI) to assess the relative left-right size of the planum temporale (PT), a part of Wernicke’s area thought to be associated with higher-level auditory processing. Their results found that the PT in nonstutterers was larger in the left hemisphere than in the right, but in stutterers, it was symmetrical in size in the two hemispheres or in some cases larger on the right. The researchers pointed out that a similar pattern has been found in individuals with dyslexia and specific language impairment. Additional anomalies found in participants who stutter were differences in various gyri (folds) in the speech and language areas of the brain. The authors of the study speculated that these differences in the brains of those who stutter may reflect deficits that interfere with information flow between Wernicke’s area (auditory cortex) and Broca’s (speech motor).
A follow-up study by Foundas, Bollich, Feldman, Corey, Hurley, and colleagues (2004) was launched to find evidence of behavior differences associated with the abnormal structure shown in the earlier study. These researchers found that matched groups of stutterers and controls had approximately the same PT asymmetry (64 percent had leftward asymmetry and 36 percent had rightward asymmetry in each group). However, the subgroup of stuttering participants who had rightward asymmetry of PT stuttered more severely and interestingly had a significantly greater response to the fluency-inducing condition of delayed auditory feedback. The authors attributed this finding to the fact that the PT is thought to be important in auditory processing of speech and language and perhaps in coordinating auditory feedback of speech with ongoing speech output. Delayed auditory feedback may have corrected an auditory feedback processing deficit causing stuttering in those with rightward PT asymmetry. It is not clear from the article whether the artificial delay in the feedback caused the stutterers to read more slowly than they would have otherwise. Slowing one’s speech rate is a common response to delayed auditory feedback and may have been the means by which the stuttering subjects who had greater rightward asymmetry of PT improved their fluency. It is not clear, however, whether improved fluency was the result of (a) an improved synchrony between auditory feedback (arriving slightly later in the cortex) and speech production; (b) improved coordination of the subcomponents of speech-language production because coordination would seem to be enhanced if any motor behavior is slowed down; or (c) both.
Sommer, Koch, Paulus, Weiller, and Büchel (2002) conducted a third neuroanatomical study in Germany. Using a process called diffusion tensor imaging, these investigators examined the density of white matter fiber tracts in the area of the left operculum, fibers which are thought to connect sensory, planning, and motor areas of the brain. They chose this area of the brain to study because two years earlier, other researchers studying activity in the left operculum found asynchrony in the sequencing stages of speech processing in stutterers (Salmelin, Schnitzler, Schmitz, & Freund, 2000). Sommer and colleagues (2002) indeed found what they suspected; fibers in the left operculum in stuttering participants were less dense than those in fluent participants. These results suggested that this structural difference in stutterers’ brains could be the basis for the dyssynchrony in processing found by Salmelin and colleagues (2000). These results complement the findings of Foundas and colleagues (2001) and Foundas and colleagues (2004), who also reported anomalies in this region of the brain. As we will see in more recent studies, the results of Sommers and colleagues (2002) are the harbinger of repeated findings of a deficit in pathways connecting speech planning areas with motor execution and sensory feedback areas that are critical in speech production.
Interesting structural anomalies–increased volumes of white matter–were found in right hemisphere structures in an MRI study of 10 adult stutterers and 10 controls (Jäncke, Hänggi, & Steinmetz, 2004). These larger volumes were found in the superior temporal gyrus, inferior frontal gyrus, and precentral gyrus in the right hemisphere. These researchers, as well as some others, have suggested that these enlarged structures may be compensatory in nature, representing a response to deficits in similar structures in the left hemisphere.
In a later brain imaging study using an updated version of the same methodology (voxel-based morphometry–that uses statistical techniques to measure size and density of very small areas of the brain), Beal, Gracco, Lafaille, and De Nil (2007) confirmed the Jäncke team (2004) findings and extended them. These authors provided evidence that auditory and speech production areas in both right and left hemisphere of stutterers had increased density compared to that of nonstutterers. Several specific brain areas were noted. First, the auditory cortex showed increased gray matter density in stutterers, supporting previously found differences in activation level apparent in the auditory cortical areas in stutterers. Second, there were areas of increased density of white matter in the right hemisphere (replicating Jäncke et al., 2004). Of particular interest is that these areas of increased density are in righthemisphere locations homologous to the left-hemisphere locations found to show less dense white matter tracts by Sommer and colleagues (2002). In other words, it’s as if either (a) the right-hemisphere white matter tracts originally developed in stutterers to have the functions of connecting sensory and motor functions that are typically seen in nonstutterers’ left hemispheres or (b) these right-hemisphere white matter tracts grew denser as a result of the stutterers’ inadequate left-hemisphere white matter tracts in a compensatory fashion.
Chang, Erickson, Ambrose, Hasegawa-Johnson, and Ludlow (2008) conducted an innovative neuroimaging study of recovered and persistent stuttering children (ages 9– 12) who were all initially identified as stuttering in their preschool years. The findings were quite remarkable. Both recovered and persistent stuttering children were found to have deficits in areas related to speech production. These deficits were reduced gray matter volume in the left inferior frontal lobe area (including Broca’s) and in bilateral temporal lobe areas that may be related to auditory perception of speech and possibly storage of auditory-perceptual targets for speech production. On the other hand, only persistent stuttering children (not those who had recovered from stuttering) were found to have another deficit–reduced density in white matter tracts subserving speech motor control. Neumann and Euler (2010) suggest that these white matter fibers are essentially part of the arcuate fasciculus. This bundle of fibers is thought to send phonological representations of words, developed in Wernicke’s area, to the motor planning and execution areas in Broca’s area.
Using some of the same technology (diffusion tensor imaging, an approach that studies the density of fibers by measuring how water diffuses around them) with young stutterers (average age, 18 years) compared to nonstuttering controls, Watkins, Smith, Davis, and Howell (2008) also found reduced density of white matter tracts– pathways which they describe as serving “the integration of articulatory planning and sensory feedback, and via connections with primary motor cortex, a substrate for execution of articulatory movements” (p. 50). Findings of the same structural deficits in young stutterers by these two groups of researchers, along with the evidence of reduced gray matter volume in bilateral temporal lobe areas (Chang et al., 2008) provide neurophysiological evidence for long-standing hypotheses that stuttering is related to reduced capacity for sensory-motor processing (e.g., Neilson, 1980). Watkins and colleagues (2008) note that structural deficits in circumscribed areas of one hemisphere rarely lead to persistent speech or language disorders, so it is not surprising that their findings reflect reduced density of white matter tracts in several areas of both hemispheres. As we will see in the next section, these abnormalities in white matter tracts underlie functional abnormalities in neural processing related to speech and language production.
Researchers recently conducted a replication of the research described in the preceding paragraphs (Cykowski, Fox, Ingham, Ingham, & Robin, 2010). This group carried out diffusion tensor imaging of white matter tracts in sensorimotor integration and speech planning areas of the brain. They studied the directional properties of the axons in these pathways and found them less diffusive parallel to the axons and more diffusive perpendicular to the axons compared to nonstutterers. In other words, these nerve bundles appear to be less efficient at carrying information along the pathways.
Using more stringent statistical analysis techniques than earlier studies, Cykowski and colleagues (2010) found that the anomalies in stutterers were restricted to lefthemisphere white matter tracts similar to those originally identified by Sommer and colleagues (2002), rather than more widespread white matter tract anomalies seen in later studies. Cykowski and colleagues (2010) also found that the neural pathways affected were not the arcuate fasciculus, as some have suggested, but instead the third division of the superior longitudinal fasciculus (SLF III), which connects speech output planning areas of the ventral frontal cortex with sensorimotor integration areas of the inferior parietal lobe. They speculated that the etiology of the less efficient transport structure of the white matter tracts in individuals who stutter may be a result of delayed myelination of nerves in these pathways.
This explanation is essentially a reprise of Karlin’s (1947) hypothesis that “the basic cause for stuttering is a delay in the myelination of the cortical areas in the brain concerned with speech” (p. 319). Karlin goes on to suggest that the myelin sheath surrounding nerves insulates them, and incomplete myelination results in slower transmission in nerves and in actions that are less precise and less coordinated. He further points out that myelination occurs earlier in girls than in boys (Flechsig, 1927), significant given the greater persistence of stuttering in boys.
It seems to me that an important effect of delayed myelination in these bidirectional nerve pathways may be that without adequate insulation provided by the myelin sheath, these pathways may be vulnerable to “cross talk” from high levels of activity in emotion and language processes (see Fig. 2.4 in Chapter 2). Vulnerability of poorly myelinated fibers in SLF III to interference by emotional activity seems possible given the evidence that the left frontal cortex is active for positive emotional arousal and the evidence reported by Johnson, Walden, Conture, and Karrass (2010) that stuttering may increase during positive emotional arousal. Furthermore, emotion-related disorders are reported to be associated with decreased myelination in the superior longitudinal fasciculus (Pavuluri & Passaroti, 2008).
In contrast to the effects of emotion (creating “noise” in the transmission lines), the effects of language processing may be more a phenomenon of overwhelming the transmission lines because so much information must be passed back and forth at such high speeds (e.g., longer sentences are spoken more quickly). Karlin (1947) points out that stuttering first appears when “sentence formation and flow of language has become more fully developed” (p. 319), implying that this increased demand of more developed language is too much for the pathways that are not fully myelinated, and thus stuttering results. Cykowski and colleagues (2010) also suggest that speechlanguage demands (such as when producing low-frequency or more complex words that require more careful monitoring) can put enough stress on the unmyelinated fiber tracts to provoke stuttering.
These findings indicate that stutterers’ left-hemisphere structures for speech and language appear to have deficits or delays in development. This may prompt the use of homologous right-hemisphere structures, which themselves may not be as suited for rapid speech production as left-hemisphere structures (Geschwind & Galaburda, 1985). This results in stuttering, especially under the stress of increasing language demands as well as interference from nearby right hemisphere centers for emotion.
Brain Function Differences in People Who Stutter
Electroencephalographic Studies
Motivated by a theory that a lack of dominance of one cerebral hemisphere over the other caused stuttering (Orton, 1927; Orton & Travis, 1929; Travis, 1931), Travis and his students used electroencephalographs (EEGs) to measure brain waves in stuttering and nonstuttering subjects. Their hope was to find proof that the brains of stutterers didn’t show the normal left-hemisphere dominance during speech and that this resulted in mistimed signals to muscles of the speech production system. EEG studies are carried out by pasting electrodes on the surface of the scalp to measure the electrical activity of brain. This procedure, like any assessment of the brain, was fraught with methodological quandaries (Bloodstein, 1995). How faithfully would electrical activity on the scalp plumb the activity of brain cells several centimeters below? How do we know that the electrical activity recorded isn’t created by muscle contractions during speech or even the result of the subject blinking her eyes or wiggling her nose? How do we know which part of the brain is active when we see the squiggles on the chart paper that represent electrical impulses? These uncertainties make it likely that the EEG studies by different scientists in different laboratories will produce widely different findings due to different methodologies and different interpretations of the data.
Despite these problems, many EEG studies of stutterers and nonstutterers have been conducted, and some interesting findings have turned up. As with other experimental results, you should be cautious about accepting the results as final proof. Several EEG studies supported the notion that stutterers’ brains functioned differently, although other studies did not. Studies by Travis and Knott (1937); Douglass (1943); Zimmermann and Knott (1974); Ponsford, Brown, Marsh, and Travis (1975); many by Moore and his colleagues (e.g., Moore & Haynes, 1980); and a study by Boberg, Yeudall, Schopflocher, and Bo-Lassen (1983) showed in different ways that individuals who stutter tended to have more activity on the right side of the brain during speech and especially during stuttering than did nonstutterers. This activity seemed to involve structures in the right hemisphere in a similar location as those in the left hemisphere that control speech and language (sometimes referred to as “homologous” areas). Although this was not exactly what Orton and Travis had predicted, it was close. Whereas the Orton-Travis theory hypothesized that stutterers lacked hemispheric dominance, findings from these EEG studies suggested that rather than lacking dominance, stutterers may be more likely to have a right-hemisphere dominance for speech and language (whereas nonstutterers generally have left hemisphere dominance).
Cerebral Blood Flow Studies
In the 1970s and 1980s, researchers developed new technology that was more precise than EEG in detecting where brain activity was occurring by measuring the amount of blood flowing to those areas. Cerebral blood flow (CBF) is usually detected by injecting a radioactive tracer into the bloodstream and taking the equivalent of x-ray pictures of the amount of radioactivity given off. The greater the amount of neural activity in an area, the greater the blood flow in that area and the greater the amount of radioactivity given off.
Interpretation of CBF and other brain imaging studies must take into account many different variables that can influence the results (Ingham, 2001). Two such variables are the spatial and temporal resolution of each technology. In other words, how accurately can CBF determine exactly where the activity is occurring and when it’s occurring? Early studies could only observe areas of the brain in a relatively general way, but improved technologies have allowed researchers to differentiate the activity of different areas in more detail and reveal interactions among brain areas. Besides the problem of how good the resolution is, other variables that can influence outcomes of brain imaging studies are the gender of the participants, the severity of their stuttering, whether or not they’ve had speech therapy, what tasks the participants perform in the study, and techniques used to analyze the data (Lauter, 1995, 1997). Keep these factors in mind as you read about the research in this area.
Wood, Stump, McKeehan, Sheldon, and Proctor (1980) published the first study of CBF in stuttering, using only two participants. They found greater activity in the righthemisphere region corresponding to Broca’s area than in Broca’s area (left hemisphere) itself during stuttering before treatment with the drug haloperidol. After two weeks of treatment with haloperidol, both participants showed that the greater activity had shifted from right- to left-hemisphere speech areas. The second CBF study of stuttering was not published until 11 years later. Pool, Devous, Freeman, Watson, and Finitzo (1991) studied the brains of 20 adult stutterers using single photon emission computed tomography, an improved technology that enabled scientists to view the brain from multiple angles and obtain better images of what was going on. The principal finding from this study was that the stuttering group showed less lefthemispheric dominance compared to controls in areas that are believed to be associated with language processing (middle temporal lobe), speech motor control (inferior frontal lobe), and motor initiation (anterior cingulate).
Positron Emission Tomography Studies and Beyond
Four years later, another CBF study took advantage of a new brain imaging tool, positron emission tomography, which allowed researchers to make more accurate inferences about where increased blood flow was occurring in the brain. A large team of researchers studied the brains of four adults who stuttered and a matched group of control participants in two conditions: reading aloud alone and in unison with someone else called “choral reading” (Wu, Maguire, Riley, Fallon, & LaCasse et al., 1995). As you may remember from Chapter 1, when individuals who stutter read aloud along with someone else, they become fluent. These findings suggested that two important speech and language areas of the brain–Broca’s area and Wernicke’s area, both in the left hemisphere (see Fig. 2.3)–showed decreased activity (compared to their normalspeaking controls) when participants were stuttering, compared to when they were fluent during choral reading.
Broca’s area is thought to be the major cortical area responsible for organizing and executing speech motor output. Wernicke’s area, on the other hand, is vital for speech and language comprehension but may also be involved in speech output because it appears to be the storehouse for the sounds that form words–the phonological representations that are called upon before the motor commands are given.
In November 1995, four different research groups presented their findings at the annual convention of the American Speech-Language-Hearing Association in Orlando, Florida (De Nil, Kroll, Houle, Ludlow, Braun, & Ingham et al., 1995; Ingham & Fox, 1995; Wu, Maguire, Riley, Fallon, & LaCasse et al., 1995). As the presentations were given, the excitement in the room was palpable because so many of their findings were similar, although the groups were working entirely independently. Here at last was clear evidence that the brains of people who stutter worked differently than those of nonstutterers. Years of previous speculation and studies suggesting anomalous cerebral dominance, inadequate laterality, auditory processing problems, and language dysfunction in stuttering seemed to be confirmed. These findings and others are reviewed below, organized by the types of anomalies they suggest.
Brain Overactivation During Stuttering
Many studies have shown that certain areas of the brain show higher levels of activation in those who stutter than in controls. Sometimes this is present during stuttering, but it has also been shown to occur in stutterers’ fluent speech. Researchers have suggested that some of the overactivations may be important etiological factors, while others may be compensatory.
Overactivation of Right-Hemisphere Cortical Areas During Stuttering
A common finding by several of the brain research teams in 1995 and afterward is that individuals who stutter demonstrate high levels of activity in the right hemisphere when they are speaking, especially when stuttering, as illustrated in Figure 2.5.
The focus of this activity is greatest in right-hemisphere structures that are homologous to those in the left hemisphere used by normal speakers (Braun, Varga, Stager, Schulz, Selbie, & Maisog et al., 1997; De Nil, Kroll, Kapur,& Houle, 2000; Fox, Ingham, Ingham, Hirsch, & Downs et al., 1996; Fox, Ingham, Ingham, Zamarripa, Xiong, & Lancaster, 2000). One of those areas is called the right frontal operculum (see Fig. 2.3) and is in the same location in the right hemisphere that Broca’s area is in the left hemisphere (Fox, 2003). Broca’s area is thought to be active in planning the phonetic structure of an utterance to be spoken (Kent, 1997). Another area in the right hemisphere commonly found to be active during stuttering is the right insula (Fox, 2003). In the left hemisphere, the insula may function as a connection between Wernicke’s area (which may be important for phonological representations of words and auditory monitoring of one’s own speech) and Broca’s area (Ingham, Ingham, Finn, & Fox, 2003).
A meta-analysis of many studies that compared stutterers and controls (Brown,
Ingham, Ingham, Laird, & Fox, 2005) confirmed that a major difference between these groups was a general overactivation of right-sided areas that are homologous to leftsided areas active for speech production: the frontal operculum, Rolandic operculum, and anterior insula. It should be noted that these researchers also observed that a common finding was overactivation in left-hemisphere areas related to motor control of speech, perhaps as a result of the extra effort required to speak.
Researchers have considered two possible explanations for the overactivation of right-hemisphere structures during stuttering. One is that during embryonic development, the right side of the brain becomes “wired” to be the primary speech and language area (e.g., Geschwind & Galaburda, 1985). This may result in some difficulty speaking because right-hemisphere structures are not generally suited for the rapid processing of signals required for speech (such as the quick transitions in many consonant-vowel transitions). When a child with this right-hemisphere “wiring” for speech develops language beyond the single-word stage, stuttering may emerge as he tries to produce multiword utterances at the typically fast speech rates used for longer sentences (Kent, 1984). A second hypothesis is that the child who stutters initially tried to use left-hemisphere regions for speech and language, but the neural networks for speech and language failed to function adequately and resulted in stuttering. Only then did the child’s brain begin to use right-hemisphere structures in a compensatory way to try to achieve more normal speech, similar to the way in which some individuals with aphasia use right-hemisphere structures to compensate for damaged areas in the left hemisphere (e.g., Sommer et al., 2002; Weiller, Isensee, Rijntjes, Huber, Müller et al., 1995).
Several researchers have provided evidence in favor of the second (compensation) hypothesis. Braun and colleagues (1997) found that activations of right-hemisphere sensory areas were negatively correlated with stuttering; that is, these regions became more active as speech became more fluent. Moreover, researchers in Germany (Neumann et al., 2003) found that right-hemisphere activations were greater in participants who stuttered moderately compared to those who stuttered severely, suggesting that right-hemisphere activity may indeed be a way in which individuals could partially overcome dysfunctions in the left-hemisphere areas. In other words, the moderate stutterers used more compensatory right-hemisphere activity to reduce the severity of their stuttering.
It is possible, of course, that both hypotheses are correct–that some individuals develop right-hemisphere processing for speech and language before they begin to stutter, and others develop right-hemisphere processing after they begin to stutter, as a compensatory response. Still others may in fact process speech and language in both hemispheres simultaneously. Each of these options is probably inefficient and may create the dyssynchrony in processing that results in stuttering.
Overactivation in Midbrain Areas
A number of researchers have reported unusually high levels of activity in midbrain structures that, via pathways to the cerebral cortex, may influence speech movements. Specifically, some structures of the basal ganglia have been shown to be overactive in those who stutter (e.g., substantia nigra, subthalamic nucleus, red nucleus, globus pallidus) (Fox et al., 1996; Watkins et al., 2008). This midbrain area is important in speech motor activity, via loops that send signals from the cortex to the basal ganglia, which then send “go” or “no go” signals to the supplementary motor area (SMA), which is responsible for initiating movement. Excess activity in the basal ganglia associated with stuttering could possibly result in inhibitory signals sent to SMA preventing the initiation of speech movements. Another perspective, given by Alm (2004), is that the excess right-hemisphere SMA activity during stuttering reported by Fox and colleagues (1996) may be the result of inadequate left-hemisphere SMA activity due to basal ganglia dysfunction in stutterers.
Underactive Brain Areas in Stuttering
Several areas of the brain, including both motor and sensory centers, have been found to be underactive in stuttering. Because speech motor control involves motor and sensory integration, it is not surprising that this is sometimes reported.
Underactivity in Speech Motor Areas
Watkins and colleagues (2008) found that during speech production, stutterers showed decreased activity compared to controls in areas related to using sensory and motor information and planning sequential movements: ventral premotor, Rolandic opercular, and sensorimotor cortex on both sides of the brain. This underactivity was in the same area of the brain as the structural differences found by Watkins and colleagues (2008)–less dense white matter tracts connecting articulatory planning and sensory feedback areas. Thus, structural differences seem to result in functional deficits that presumably interfere with the smooth flow of speech. The researchers point out that the fact that the decreased activity is present on both sides of the brain and is widespread may account for the fact that, unlike in cases of unilateral and limited brain lesions, these individuals’ brains were unable to develop work-arounds to compensate for the problem.
Underactivity in Auditory Areas
Many brain imaging studies of stuttering have shown a lack of activity in the superior temporal lobe, including auditory association areas and Wernicke’s area (Braun et al., 1997; De Nil, Kroll, Lafaille, & Houle, 2003; Fox et al., 1996; Fox et al., 2000). The meta-analysis by Brown and colleagues (2005) mentioned earlier indicated that a common finding among several of these studies was that auditory areas in both hemispheres were deactivated, suggesting that stutterers’ mechanisms for guiding their speech by self-hearing were not functioning properly. More recently, Watkins et al. (2008) found in their study of adolescent and young adult stutterers that underactivity was notable in Heschl’s gyrus, a part of the superior temporal lobe that is important for processing speech sounds. Another imaging study (Salmelin, Schnitzler, Schmitz, Jancke, Witte, & Freund, 1998) found that stutterers have a reversal of the normal pattern of activation of the left and right auditory cortices during stuttering. Evidence of auditory dysfunction during stuttering is especially pertinent in light of the many studies that have shown that stutterers may have difficulty performing auditory processing tasks (e.g., Barasch, Guitar, McCauley, & Absher, 2000; Molt, 1998) and that fluency can be induced by changing the way stutterers hear their own speech (e.g., Brayton & Conture, 1978; Howell, El-Yaniv, & Powell, 1987).
How does auditory self-monitoring affect fluency? It may provide a stimulus to synchronize or integrate the sequence of activities that run in parallel when a speaker decides what she will say, selects the linguistic elements for it, and executes the utterance. Thus, the asynchrony or timing disturbance that many researchers see as the basis of stuttering (e.g., Perkins, Kent, & Curlee, 1991; Van Riper, 1982) may be caused by a paucity of signals that synchronize the sequence for speech output. Therapies (e.g., Van Riper, 1973) that emphasize proprioception may be giving the client another feedback modality to use for timing; therapies that focus on the use of slow speech, gentle onsets, and light articulatory contacts may develop the client’s auditory and proprioceptive monitoring of speech.
Other functions besides monitoring one’s own speech may also reside in the deactivated regions of the superior temporal lobe. For example, Wernicke’s area may be important for storing the phonological representations of words (Caplan, 1987; Paulesu, Frith, & Frackowiak, 1993). Activation of this region of the brain, therefore, may be a key stage in phonological planning for speech production. Lack of activation during stuttering may reflect a deficit in the sequence of phonological selection, phonetic planning, and motor execution.
Research has assessed how well individuals who stutter can process auditory signals in various parts of the brain. A number of studies have used the Synthetic Sentence Identification/Ipsilateral Competing Message test (SSI-ICM) to compare stutterers and nonstutterers. This test requires participants to identify words in a nonsense phrase (such as “small boat with a picture has become”) when competing noise is presented in the same ear (right on top of the nonsense phrase). Three studies using this test found that stutterers performed worse than normal participants (Hall & Jerger, 1978; Molt & Guilford, 1979; Toscher & Rupp, 1978). The same test was given to two groups of normal speakers–those who were judged to be more fluent and those who were judged to be less fluent. The more fluent normal speakers performed significantly better than the less fluent normal speakers (Wynne & Boehmler, 1982). This finding suggests that stuttering and normal disfluencies are not entirely different phenomena; both may be associated with some difficulty in central auditory processing. This difficulty may give rise to disfluencies in many children, only some of whom develop chronic stuttering. Some other factor in addition to an auditory processing difficulty may be necessary for high levels of normal disfluency to become stuttering.
In contrast to the above studies, Hannley and Dorman (1982) found no differences between stutterers and nonstutterers on the SSI-ICM, but the stutterers in their study had all recently completed a treatment program. This finding is intriguing in light of evidence from brain imaging studies that individuals who stutter who had demonstrated an absence of activity before treatment in the left auditory cortex showed normal levels of activity immediately after treatment (De Nil et al., 2003; Ingham, 2003; Neumann et al., 2003; Stager, Jeffries, & Braun, 2003).
Another tool for assessing central auditory processing is the Masking Level
Difference (MLD) test, which requires listeners to detect the onset and offset of a tone in the presence of a masking noise. When masking noise is played in the same ear as the tone, there are fewer cues for listeners to use in “filtering” the tone from the masking noise. Listeners must use very subtle temporal cues to detect the tone; under these conditions, persons who stutter perform more poorly than groups of nonstutterers (Liebetrau & Daly, 1981; Kramer, Green, & Guitar, 1987). These results may be interpreted to support the outcome of the SSI studies because both tests require the participants to use temporal information–in one case (SSI), the information is in rapidly changing format frequencies (characteristic acoustic information) needed to identify words, and in the other case (MLD), the information is the onset and offset of a tone in masking.
Two other studies of central auditory processing tested the hypothesis that people who stutter have difficulty resolving temporal differences. Herndon (1966) found that stutterers were poorer than nonstutterers at distinguishing which of two brief tones was longer. Barasch and colleagues (2000) administered the Duration Pattern Sequence (DPS) test, which involves judging the relative lengths of three tones, and another measure in which subjects estimated the protensity (i.e., durations) of tones and silent intervals. These tests failed to distinguish between the stuttering and nonstuttering participants as groups, but they showed that less fluent participants in each group scored more poorly on the DPS than those who were more fluent. In addition, the more disfluent subjects in both groups judged temporal intervals to be longer than did less disfluent subjects. This finding is evidence that there may be a connection between normal disfluency and stuttering. The authors speculated that one of the possible connections between increased disfluency and longer protensity estimates is the effect of fear and anxiety on speaking. It has been suggested that fear and anxiety affect temporal processing (Fraisse, 1963) and that anomalies in temporal processing may be an underlying cause of both stuttering (Kent, 1984) and high levels of normal disfluency (Wynne & Boehmler, 1982). Given the evidence that stutterers are no more anxious than nonstutterers (Guitar, 1998), one might ask–why should stutterers’ temporal processing for speech be any more susceptible to negative emotions than temporal processing in nonstutterers? The answer may be found in evidence that negative emotions such as fear and anxiety appear to be regulated by the right hemisphere along with evidence that stutterers process speech in widely distributed areas of the right hemisphere. Thus, the close proximity of areas regulating negative emotions and speech production may allow “cross talk” between these areas to interfere with speech, producing disfluency. It remains to be seen if highly normally disfluent speakers also use right-hemisphere structures for speech production.
As described in the earlier section on brain structure differences in stutterers, Foundas and colleagues (2004) looked for differences in auditory processing that might reflect the differences in the structure of a part of the auditory temporal cortex (PT). They did indeed find an auditory processing difference in a subgroup (36 percent) of stutterers who had an atypical rightward asymmetry in this part of the auditory cortex. As you may remember from the earlier discussion of this work, the difference was that stutterers with atypical auditory cortices became more fluent when speaking while listening to delayed auditory feedback. This suggests that for some stutterers, speech production timing and coordination difficulties may originate in dyssynchronies in central processing of auditory feedback.
Brain Electrical Potentials Reflecting Auditory Processing
Studies of electrical brain activity in response to auditory stimuli have provided further evidence that auditory processing is abnormal in individuals who stutter. Studies by Hood (1987) and Dietrich, Barry, and Parker (1995) reflecting both subcortical and cortical activity have found group differences between stutterers and nonstutterers. However, the first study found stutterers’ responses to be slower, and the second found them to be faster than those of nonstutterers. A study by Molt (1998) is more relevant to the question raised by brain imaging studies of whether persons who stutter have a deficit in the left auditory cortex. Molt found that stutterers had longer latencies and lower amplitudes of brain waves in the cortex when they were asked to make decisions about semantic incongruencies in sentences to which they listened when compared to nonstutterers.
In a study that combined behavioral and brain electrical activity measures related to auditory processing of nonlinguistic material (tones), Hampton and Weber-Fox (2008) showed that while most of a group of 11 stutterers performed like a matched group of 11 nonstutterers, there was a subgroup of three stutterers who performed much more poorly than the nonstutterers on behavioral measures (accuracy and speed of identification of tones) and showed abnormalities in the brain electrical activity measures (early “P300” responses, positive-going peaks at a point 300 msec after the stimuli). This finding is reminiscent of the subgroup of stutterers described in the Foundas et al. (2004) study with an anomalous rightward asymmetry in the PT who showed a greater improvement in fluency while speaking under delayed auditory feedback than did those stutterers with more typical PT asymmetry. In other words, not all individuals who stutter may have auditory processing anomalies. A small group may have both structural and functional deviations from the norm, and this group may benefit from therapeutic approaches that help them compensate for these deviations.
Dichotic Listening Tests
More support for the notion that stutterers have abnormal auditory processing comes from speech perception studies. A number of experiments found that many persons who stutter do not show the typical right-ear advantage that nonstutterers do, evidence that people who stutter do not have left-hemisphere dominance for language (Blood, 1985; Curry & Gregory, 1969; Davenport, 1977; Liebetrau & Daly, 1981; Sommers, Brady, & Moore, 1975). Some dichotic studies, however, found no differences between stutterers and nonstutterers (Dorman & Porter, 1975; Pinsky & McAdam, 1980; Slorach & Noehr, 1973). Other studies found no significant group differences but found that fewer stutterers than nonstutterers showed the expected right-ear advantage (Brady & Berson, 1975; Quinn, 1972; Rosenfield & Goodglass, 1980; Strong, 1977) or found that the magnitude of difference between ears was greater for nonstutterers (Blood & Blood, 1989).
Looking over the myriad of studies that have been conducted, it appears that the more linguistically complex the stimulus (e.g., words versus syllables), the more likely that differences between stutterers and nonstutterers would be found. Thus, stutterers’ differences appear to be related to their linguistic processing abilities; they appear more likely to process linguistically meaningful stimuli in their right hemispheres.
Like much research in this area, dichotic testing of stutterers also suggests there might be group differences or at least subgroups of stutterers who differ from nonstutterers on a critical dimension. Examples of such subgroups might be those with more severe stuttering (Davenport, 1977); those who show several abnormal signs on neuropsychological test batteries, suggesting an “organic” origin of stuttering; and those showing only a few abnormal signs, suggesting a “functional” origin (Liebetrau & Daly, 1981). Perhaps the best example, however, is the stuttering subgroup mentioned in the previous section–individuals who show structural and functional abnormalities in sensory–especially auditory–processing.
Auditory, Tactile, and Proprioceptive Feedback
A connection between stutterers’ brain anomalies in auditory processing and speech performance was suggested by Stromsta (1957, 1972, 1986). His research on auditory feedback led him to suggest that stutterers’ abnormal brain rhythms interfere with the smooth integration of auditory feedback from the speaker’s own voice and his subsequent speech output. The result, he suggested, is a combination of a stoppage of phonation and an improper coarticulation of sounds (Stromsta, 1986). This proposal foreshadows the results of recent brain imaging studies reviewed earlier. For example, Stager, Jeffries, and Braun (2003) suggested that brain scans during conditions that promoted fluency showed increased activity in stutterers’ central auditory cortex, reflecting “more effective coupling of auditory and motor systems” (p. 334) so that auditory feedback could facilitate the sequencing of speech motor output. And of course, the results of the studies by Foundas and colleagues (2004) showing atypical morphology in some stutterers’ auditory temporal cortex also suggest that problems in processing auditory feedback may underlie stuttering.
The few studies that have been conducted of other feedback systems besides auditory also show some deficits, but the results are mixed. Baker (1967) found that people who stutter performed more poorly than nonstutterers on tests of oral sensation. However, this finding was not replicated by Jensen, Sheehan, Williams, and LaPointe (1975). On a test that required subjects to match spatially ordered visual patterns with temporally ordered auditory patterns, Cohen and Hanson (1975) found that individuals who stutter performed more poorly than those who didn’t. Chuang, Fromm, Ewanowski, and Abbs (1980) evaluated stutterers’ abilities to make the smallest movements possible with their jaws and tongues. The stuttering group had significantly larger “difference limens” (smallest detectable difference) with or without the assistance of visual feedback for such movements. This means that they did not have the degree of fine sensory-motor control of the jaw and tongue that the nonstuttering group did. De Nil and Abbs (1991) followed up on this study and demonstrated that stutterers had less sensorimotor control for minimal movements with their jaws, lips, and tongues (but not finger movements) compared to nonstutterers, when using only kinesthetic feedback. There were no differences between the groups when using visual feedback.
Together with the findings about the auditory system, these studies may indicate that as a group, individuals who stutter have some difficulty using auditory, touch, and movement information to control speech. But Namasivayam, van Lieshout, McIlroy, and De Nil (2009) provided evidence that stutterers are as capable as nonstutterers in using sensory feedback to stabilize speech motor control in the face of experimental perturbations (masking noise and tendon vibration). If they are as good as nonstutterers at using sensory feedback to nullify the effects of external interference with speech movements, stutterers may be using sensory feedback to also nullify the effects of internal interference with speech (i.e., their innately poorer speech motor skills). In other words, when stutterers are fluent, as they naturally are some of the time, they may be achieving this fluency by making extra use of sensory feedback (van Lieshout, Hulstijn, & Peters, 2004). In summary, this literature is full of conflicting findings, and it is not clear whether all stutterers have deficits in sensory feedback or whether there is only a subgroup that shows these deficits.
The first experiments on people who stutter found that they were slower than nonstutterers in initiating and terminating a vowel in response to a buzzer (Adams & Hayden, 1976; Starkweather, Hirschman, & Tannenbaum, 1976). Later experiments showed that stutterers were slower than nonstutterers in reacting with respiratory (exhalation) and articulatory movements (lip closing) (McFarlane & Prins, 1978; Watson & Alfonso, 1987). They were also slower whether they were responding to auditory or visual signals (e.g., Cross & Cooke, 1979). Children who stutter were also found to have slower reaction times in studies by Cross and Luper (1979, 1983), Cullinan and Springer (1980), Maske-Cash and Curlee (1995), and Till, Reich, Dickey, and Sieber (1983).
Although not all studies showed group differences, De Nil (1995) pointed out that about 75 percent of the 44 voice reaction time studies that he reviewed found that people who stutter were significantly slower than people who don’t stutter and that most of the other studies showed trends in that direction. He further noted that when investigators used linguistically meaningful stimuli to test reaction times (words or sentences, rather than isolated sounds), 80 percent of the studies found significant differences between stutterers and nonstutterers. These findings are likely to be related to the evidence from brain imaging studies indicating that individuals who stutter have anomalies in areas used for sensorimotor processing of speech and language. Not surprisingly, these anomalies may affect sensorimotor reaction times on nonspeech tasks, but are most evident in tasks requiring linguistic processing.
Fluent Speech
Acoustic studies have demonstrated that, on average, stutterers have longer vowel durations, slower transitions between consonants and vowels, and delayed onsets of voicing after voiceless consonants even when speaking fluently (Colcord & Adams, 1979; DiSimoni, 1974; Hillman & Gilbert, 1977; Starkweather & Myers, 1979). The findings of these acoustic studies have been supported by “kinematic” research, which has measured the movements of speakers’ speech structures (e.g., Alfonso, Story, &
Watson, 1987; Zimmermann, 1980). As a group, stutterers tend to move their lips and jaws more slowly, even during fluent speech, than do nonstutterers (e.g., Zimmermann, 1980). Kinematic research has also shown that some stutterers demonstrate abnormal sequencing of articulator movement onsets and velocities (Caruso, Abbs, & Gracco, 1988). Other kinematic studies, however, have not found group differences or have found them only in stutterers who had recently undergone therapy (e.g., McClean, Kroll, & Loftus, 1990), which may have taught them to speak more slowly.
So what does this mean? Why might many stutterers speak more slowly or use different sequences of articulatory movements than nonstuttering speakers even when they are fluent? Some researchers think that these findings reflect delays or other dysfunctions in processing incoming and outgoing signals. Stutterers may be unable to process neural signals fast enough to make the rapid, precise movements of normal conversational speech, especially when they are under the stress of planning a complex sentence or competing with other talkers. Their delays in voicing onset, slower transitions, and abnormal sequencing during fluent speech may just reflect a slower mechanism working at its normal rate. A different view, suggested by more skeptical researchers, is that such differences simply reflect the way stutterers have learned to talk to avoid stuttering, either on their own or as a result of therapy, and this way of speaking keeps them fluent even with an inefficient speech motor system.
Another interpretation of these slower movements is that they are the result of heightened tension in muscles having antagonistic functions for speech production (Starkweather, 1987). For example, increased tension in muscles that move a structure forward (agonists), as well as muscles that hold it back (antagonists), would make movement of that structure considerably slower. Imagine two people pulling a rope in opposite directions. Even if one were stronger, that person would make slow progress in her direction because the other, weaker person would create a drag. The slowed movements of stutterers’ speech structures would account for not only slower reaction times but also the longer movement durations in their fluent speech.
Findings from a number of studies support this view. They have shown that stutterers co-contract agonist and antagonist muscles of both the laryngeal (Freeman & Ushijima, 1975; Shapiro, 1980) and articulatory (Guitar, Guitar, Neilson, O’Dwyer, & Andrews, 1988) muscle groups during stuttering. These studies, like Starkweather’s (1987) review, have noted that such co-contraction of agonist and antagonist muscles appears even in some of the apparently fluent speech of stutterers. This finding has led many researchers to posit that stuttering is not an “all-or-nothing” event (Adams & Runyan, 1981; Bloodstein, 1987). Sometimes, stutterers may speak freely, without a trace of excess tension. At other times, they may have excess tension that isn’t heard by listeners as stuttering. At still other times, muscle tension may be so great that both listeners and the person who stutters are acutely aware of stuttering. This continuum of fluency reflects the subjective impression of many stutterers, including me.
Nonspeech Motor Control
In a study of both sequential finger movements and sequential counting aloud fluently, Borden (1983) found that severe stutterers, but not mild stutterers, were slower than nonstutterers in executing both finger movement and speech tasks. Thus, severe stutterers may have substantial deficits in certain sensory-motor tasks, but mild stutterers only slight deficits, and these may require special task conditions to be revealed.
Webster (1993a) developed a finger movement task 10 years later in which participants tapped four numbered keys in a predetermined sequence. To make the task somewhat like speech, participants were assigned a novel sequence of keys at the beginning of each trial (3-2-4-1 or 4-1-2-3, etc.). In both timed and untimed tests, stutterers made more errors sequencing and were slower initiating the task but were comparable to nonstutterers in execution time (once the movement was started). Unlike Borden’s study, no effort was made to analyze the results by subjects’ stuttering severity. Webster thought that these results suggested that stutterers may have difficulty in “response planning, organization, and initiation” (Webster, 1993b, p. 84) of novel sequences of movements.
To answer the question of why this difficulty may be present only intermittently in individuals who stutter (after all, they have a great deal of fluent speech), Webster (1993b) postulated what others (Cross, Sweet, & Bates, 1985; Curlee, 1993; Peters & Guitar, 1991) have also considered that at times, such as under stress, there is “cross talk” or interference between the hemispheres. Specifically, activity in the right hemisphere interferes with sequential movement control in the left hemisphere.
To test this, Webster used a task in which participants performed sequential finger tapping with the right hand while turning a knob with the left hand, in response to an auditory signal. If his hypothesis were true, the stuttering group’s left hemisphere– controlled finger tapping would be vulnerable to interference by the right hemisphere– controlled knob turning. Indeed, the stuttering group’s performance was significantly poorer than that of the nonstuttering group on both tasks.
Webster (1997) then wondered if interference of left-hemisphere sequential movement control mechanisms might be the result of stutterers’ inability to focus attention on the left-hemisphere task and ignore interference from a competing source, whether from the right or left hemisphere. To test this, Webster used a procedure developed to investigate attention focus in right- and left-handed people. Participants were required to tap twice with one hand for every tap they made with the other hand. They were tested with the right-hand double tapping and the left single tapping, as well as vice versa. The nonstutterers were able to perform the task significantly better when they tapped twice with the right hand and once with the left; however, stutterers and nonstuttering left-handers performed the task equally well with either hand doing the double tapping. Webster interpreted this outcome as suggesting that stutterers and left-handed nonstutterers did not have the ability to focus predominantly on the left hemisphere, but had equal focus on both, making their left hemispheres vulnerable to interference from other activities. Webster’s model of stuttering, derived from these experiments, postulates that individuals who stutter are unable to protect the integrity of speech production centers from interference or “cross talk” from emotions or other ongoing processes. It is not clear why then, all left-handers don’t stutter. I would presume that this model proposes that stutterers have both a deficit in the sequencing of processing underlying speech production as well as an inability to focus on the left hemisphere.
Following up on Webster’s studies, Subramanian and Yairi (2006) experimented with a version of the finger-tapping task, using not only stutterers and controls but a third group of “high-risk” participants. All participants were right-handed. The highrisk individuals were parents or siblings of the stutterers–of special interest because they were thought to be carriers of genetic material that could create stuttering except for the absence of some critical factor that would presumably cause them to stutter, too. All participants tapped in several conditions, including tapping at a comfortable rate and at a fast rate. Among the findings of this study was evidence that in the comfortable rate condition, the stuttering group and the high-risk group were slower than the control groups. However, in the fast rate condition, the stuttering group tapped faster than either of the other groups, both with their right hands and with their left hands. Along with their high rate, the stuttering group also had higher variability in tapping rate than either group.
Of great interest was the finding that in this fast rate condition, the high-risk group had slower tapping rates than either the stuttering group or the control group but had relatively low variability. The authors speculated that these findings may reflect that those in the stuttering and high-risk groups have motor systems that operate best at slow rates. When there is pressure to operate at a fast rate, the high-risk group is able to control their rate and maintain stability (low variability). But under this pressure, the stuttering group taps very rapidly, pushing their motor systems beyond their optimal operating speed, causing them to become unstable (high variability).
It is not too great a leap to imagine that individuals who stutter speak more rapidly than is optimal for their speech planning and execution systems, thus becoming disfluent. It is also possible that close relatives of individuals who stutter who may have inherited a predisposition for stuttering don’t develop stuttering because they are able speak at slower speech rates, more appropriate for the limitations of the speech motor production system. It is interesting that this is the same conclusion drawn by Kloth, Janssen, Kraaimaat, and Brutten (1995, 1998) who found that in a high-risk-forstuttering population of 93 children studied prior to onset, the 26 children who did develop stuttering spoke more rapidly than the 67 children who didn’t. However, both groups’ speech rates were within the normal range.
Subramanian and Yairi (2006) also found evidence that could support Webster’s (1997) hypothesis that stutterers suffer from an inability to focus entirely on the left hemisphere when performing motor tasks and thus may activate both hemispheres for speech, as brain imaging studies have shown (e.g., Brown et al., 2005). Subramanian and Yairi’s (2006) data showed that when participants were asked to tap simultaneously with both hands, but tap twice as fast with one hand than with the other, the stuttering and high-risk participants performed equally well with either the right or left hand tapping twice as fast. The control groups were better when they tapped twice as fast with their dominant (right) hands as with the left hands. This finding supports the speculation that the stuttering and high-risk groups did not suppress the right hemisphere and focus entirely on the left hemisphere. Another similarity to Webster’s (1997) findings was that Subramanian and Yairi’s (2006) stuttering group required more trials to learn to tap twice as fast with one hand than the other. Webster’s similar finding that stuttering subjects took longer to initiate a pattern of tapping led him to speculate that it reflects stutterers’ difficulty initiating speech movement at the beginning of an utterance, thus stuttering more at the beginnings of phrases.
Subramanian and Yairi’s (2006) finding that adults who stutter manifest greater variability in a nonspeech motor task was recently discovered in children who stutter. Olander, Smith, and Zelaznik (2010) studied 17 children who stutter and controls, ages 4 to 6, using a task that required children to clap their hands in time to a metronome and then continue to clap at that rate after the metronome was turned off. Their results indicated that although 40 percent of the children performed like the control group, 60 percent showed variability outside the range of the controls. They speculate that nonspeech motor variability may be related to speech motor control deficits and that these deficits might be predictive of recovery or persistence of stuttering.
The last studies of nonspeech motor control I will review concern the use of auditory input to control motor output. A common paradigm in these studies is to have participants track or follow the changing frequency (pitch) of a target sound with a second sound, called a “cursor.” A computer controls the pitch changes in the target’s sound and participants follow these changes by using their hands or their jaws to move a lever that changes the pitch of the cursor. Using this paradigm, Sussman and MacNeilage (1975) found that normal speakers made fewer errors tracking the target sound when the cursor tone was presented to the right ear and the target tone to the left ear. Those who stuttered, on the other hand, made equal numbers of errors whether the cursor tone was in the right ear and the target tone in the left or vice versa, suggesting that they did not have a left-hemisphere advantage for integrating auditory information with motor output as the nonstutterers did.
Researchers in Australia replicated and extended this work on tracking (Neilson, 1980; Neilson & Neilson, 1987, 1988) using both visual and auditory targets and cursors. They demonstrated that participants who stuttered were significantly poorer using auditory targets and cursors than when using visual ones. They also showed that when both stuttering and nonstuttering participants practiced the tasks beforehand, the differences between the groups were even larger than when they had not practiced. The Neilsons proposed that stutterers were slow in developing a mental auditorymotor model of the relationship between their movement of the cursor control and the resulting sound change. They further hypothesized that the basic deficit in stuttering is difficulty in forming or accessing auditory-motor models of what speech movements are needed to produce the sounds they want to make. I suspect that the Neilsons would agree that stutterers can use their auditory-motor models better in situations where the demands on their neural resources are low but have more trouble when the demands are high.
As the Neilsons were working on their experiments in Australia, researchers at Baylor College of Medicine in Texas were also studying the auditory-motor tracking abilities of persons who stutter. In a series of publications (Nudelman, Herbrich, Hoyt,
& Rosenfield, 1987; 1989; Nudelman, Herbrich, Hess, Hoyt, & Rosenfield, 1992), these researchers described experiments, which required participants to hum along with a tone that suddenly changed pitch. The pitch of the target tones was sometimes changed rapidly, sometimes slowly, while researchers measured how quickly participants could change their humming to match the changing pitch of the auditory target tone. The researchers found that stutterers were significantly slower than nonstutterers in detecting changes in the target’s frequency, suggesting that stutterers need more time to process auditory signals. Once again, we find that research has produced evidence that, as a group, people who stutter have difficulty performing auditory processing tasks–perhaps more so when the auditory information is related to vocal output. Note that this limitation might be related to findings by Cykowski and colleagues (2010) and Sommer and colleagues (2002) that brain areas used for sensory integration are not efficiently connected to motor planning/motor execution areas.
Studies of stutterer-nonstutterer differences on sensory and sensory-motor tasks, reviewed earlier in this chapter supported the notion that language factors are important in stuttering, affecting both speech and nonspeech tasks. For example, the results of dichotic listening studies and reaction time experiments indicate that there are greater stutterer-nonstutterer differences when the stimuli used are more linguistically complex and more meaningful (e.g., Curry & Gregory, 1969; De Nil, 1995). Moreover, Kleinow and Smith (2000) found that linguistic complexity affected the stability of speech motor control in stutterers. Using a measure called the spatiotemporal index that reflects the amount of variability in articulator movement when a sentence is said over and over, Kleinow and Smith demonstrated that longer and more syntactically complex sentences produced significantly more instability in stutterers’ speech production than simpler and shorter sentences. The same was not true for nonstutterers. Kleinow and Smith interpreted these findings as suggesting that stutterers’ speech production systems are more vulnerable to breakdown when language demands are high. Smith, Sadagopan, Walsh, and Weber-Fox (2010) also found increasing interarticulatory coordination breakdown as phonological complexity increased for adults who stutter but not controls.
Some researchers caution against the view that language factors are a primary causal influence in stuttering. De Nil, for example, suggests that the basic problem of stuttering may be in “sensorimotor processes involved in speech production” rather than in “cognitive language formulation processes” involved in speech production (De Nil, 2004, p. 123). He and his fellow researchers (De Nil, Kroll, & Houle, 2001) compared brain scans of stutterers and nonstutterers while they performed two tasks– one, reading single words aloud, and the other, generating verbs. While both tasks involved saying single words aloud, the verb-generation task engaged more language processing–specifically, semantic and phonological searching and encoding. To assess whether the two groups differed on the two tasks, the experimenters “subtracted” the scans of reading the words aloud from the scans of generating verbs. In other words, they wanted to see if the extra demand of generating verbs alone showed a big difference between the two groups, so they took out all activity in the verb-generating task that was also present just in reading words aloud. The subtraction indicated that there were minimal differences between the groups in the more linguistically involved task or generating verbs, although there were large differences between the groups in reading words aloud. This suggested that the added linguistic burden of having to think of verbs was not particularly more demanding on the stutterers than on the nonstutterers.
The two views (a) that language is not a primary factor in stuttering and (b) that language is a major influence in stuttering may be compatible if one assumes that the basic defect is in sensorimotor control of speech but that individuals who stutter achieve fluency by compensating for that sensorimotor defect by using additional neuronal resources, additional attention, or simplifying the task by slowing their speech rates. When language becomes more complex and/or sentences become longer, speech rate increases (Starkweather, 1981), and it would seem from evidence that there is more distributed brain activity when language is more complex that more neuronal resources and attention are devoted to the greater language demands. Thus, fewer extra resources are available to compensate for the sensorimotor defect and more stuttering results.
The notion that stutterers have a limitation on their neural resources during speech and language production is supported by the research program of Bosshardt (2006) that provides data to suggest that stutterers are poorer at tasks requiring linguistic processing while engaging in other cognitive activities. Bosshardt’s interpretation of his results describes the connection between language and stuttering by hypothesizing that individuals who stutter have speech and language planning and production systems that are less protected from interference, either by heavy processing loads within the system or by other activities in the brain. One might extend that argument to suggest that when children are first learning to use new language constructions during childhood, the neuronal resource requirements for language planning are high. Thus, children with vulnerable speech and language systems will experience interference by language planning on the smooth production of speech. Even when these children have grown older and have learned adult language, their use of long, complex sentences at rapid rates while planning the next parts of a narrative can be disrupted.
Anxiety and Autonomic Arousal
A direct connection between a momentary experience of anxiety and stuttering-like behavior in nonstutterers was suggested by an early study by Harris Hill (1954). Subjects were trained to produce a sentence describing a picture when a red light came on. After several trials, they were then given an electric shock while speaking the sentence. Subsequently, no shock was given during sentence production, but the red light was assumed to be associated with electric shock. When speaking under the anticipation of shock, subjects produced “compulsive and preservative” repetitions, prolongations, and blocks. Electromyographic sensors detected increased muscle tension during these responses. Hill reported that many responses appeared to be “indistinguishable from what is generally termed stuttering. [Moreover, the responses of several subjects] would have been classed as severe in any speech clinic” (p. 8).
Thus, it appears that even nonstutterers will show stuttering-like behaviors under threat of penalty, and some nonstutterers show severe instances of this. This study demonstrates that emotion (probably negative arousal, in this case) can cause disfluencies.
Using Bosshardt’s (2006) model for the effect of increasingly complex linguistic tasks on fluency, we might suppose that if those who stutter have, as described above, speech and language planning and production systems that are less protected from interference by other brain activities, emotion can be one of those other brain activities that can interfere with the smooth flow of speech. Hill’s finding that physical tension increases under threat of negative consequences (electric shock) would lead one to suspect that tension in stuttering also results from the threat of negative consequences. In the case of stuttering, however, those consequences may be the frustration, embarrassment, and shame that often arise when one can’t perform the simple act of saying a word. It may be significant that in Hill’s experiment, a subgroup of some subjects were much more severe than others. Perhaps this reflects individual differences in the vulnerability of the speech production system to interference. I will return to this point when I discuss temperament in the next section.
Another early study of anxiety and stuttering is of interest because of what it might tell us about emotion, speech physiology, and stuttering. Horovitz, Johnson, Pearlman, Schaffer, and Hedin (1978) looked at a phenomenon called the stapedial reflex, which had been previously shown to increase during anxiety in normal speakers. The stapedial reflex is muscle contraction in the middle ear, triggered by activation of the laryngeal nerve just prior to speaking, decreasing the loudness with which a speaker hears her own voice. The researchers found that stutterers demonstrated an increased stapedial reflex when they became more anxious (as measured by a physiological assessment of anxiety), compared to a no-anxiety condition. A group of matched nonstutterers showed no increase in stapedial reflex when their anxiety increased. Participants in both groups increased their anxiety by imagining themselves in stressful speaking situations. Although the results of this study are hard to interpret, it appears to show that an increase in anxiety in stutterers may result in changes in speech-related physiology even when only imagining some difficulty speaking. It may be relevant that the increase in stapedial reflex may have been brought about by an increase in laryngeal nerve activity. This connection between autonomic arousal and heightened laryngeal muscle activity may reflect a conditioned response that becomes part of the learning, which maintains stuttering in some individuals. I will revisit this connection between laryngeal tension and autonomic arousal when I discuss temperament and stuttering.
More recently, researchers have asked whether people who stutter are more anxious than people who don’t. To answer this question, they have used various measures of physiological arousal, such as heart rate, skin conductance, and level of cortisol (a chemical secreted by the brain under conditions of stress) in saliva secretion. Four studies (Caruso, Chodzko-Zajko, Bidinger, & Sommers, 1994; Miller, 1993; Peters & Hulstijn, 1984; Weber & Smith, 1990) found that both stutterers and nonstutterers showed high levels of autonomic arousal when they had to speak or read aloud, indicating that people who stutter are not more anxious or more nervous than people who don’t. However, several studies of autonomic arousal in individuals who stutter have reported that higher levels of arousal are associated with more stuttering (Caruso, Chodzko-Zajko, & McClowry, 1995; Miller, 1993; Weber & Smith, 1990). Although most speakers may show increased arousal when they have to speak, it may be that only the speech of those who stutter is vulnerable to breakdown (unless arousal is unusually high, as in the Hill’s 1954 experiment).
In a questionnaire study, Oyler (1992) found that adults who stutter were more emotionally sensitive than were adults who don’t stutter; however, this hypersensitivity could be the result of many years of stuttering. Greater sensitivity in children who stutter has been reported by several studies. Fowlie and Cooper (1978) reported that mothers of children who stutter viewed them as more sensitive than did mothers describing children who do not stutter. Oyler and Ramig (1995) found that parents of children who stutter rated them as more sensitive than did control parents rating nonstuttering children.
Using the concept of “difficult” temperament, which includes some aspects of sensitivity as well as restlessness and impulsiveness, both Wakaba (1998) and Embrechts and Ebben (1999) found that parents of stuttering children rated their children as having this type of temperament to a greater degree than did parents of nonstuttering children. LaSalle (1999) presented a paper indicating that, in contrast to parents of young children who don’t stutter, parents of young children who do stutter rated their children as having high frustration reactions and lack of persistence. Both of these traits have been associated with sensitive temperament (Thomas & Chess, 1977). Anderson, Pellowski, Conture, and Kelly (2003), using the Behavioral Style Questionnaire (McDevitt & Carey, 1978) found that parents of children who stutter rated their children as slower to adapt novelty compared with how parents of nonstuttering children rated their children. They related this to Kagan’s (1989, 1994) description of this personality trait as also being more shy and fearful when encountering unfamiliar events and people. More evidence of children who stutter having a more reactive temperament comes from an important study by Karrass, Walden, Conture, Graham, Arnold, and colleagues (2006). Their findings, using a scale of children’s reactions to everyday stressful situations completed by parents, suggested that when compared to nonstuttering children, children who stutter are more emotionally reactive and are less able to regulate their emotional responses. The authors speculated that these traits may make it more likely that children who stutter will react emotionally to their disfluencies, producing more disfluencies in a cycle of reactivity and increasing stuttering, which the authors call “reverberant interaction.” It should be remembered that emotional reactivity may increase laryngeal tension (e.g., Kagan, Reznick, & Snidman, 1987) and perhaps tension in other parts of the speech system, not only increasing stuttering, but making it more severe as well.
Some researchers (e.g., Kagan, 1994a, 1994b) have advocated for physiological or behavioral studies, rather than parent rating of children’s temperament. In this spirit, using a physiological measure of sensitivity, Guitar (2003) found that the acoustic startle responses of adults who stutter were significantly greater than those of adults who do not. The startle paradigm, which measures the magnitude of the eye blink in response to a burst of white noise, is believed to differentiate individuals whose nervous systems have low thresholds of arousal from those whose nervous systems require larger stimuli to react (Vrana, Spence, & Lang, 1988). Moreover, it has demonstrated differences in children who have been categorized as temperamentally inhibited and temperamentally uninhibited (Snidman & Kagan, 1994). Guitar (2003) also found substantial correlations between startle responses and scores on the nervous subscale of the Taylor-Johnson Temperament Analysis (Taylor & Morrison, 1996).
It should be noted that two later studies, Alm and Risberg (2007) and Ellis, Finan, and Ramig (2008), failed to replicate Guitar’s (2003) findings. This suggests that a sensitive temperament may not be a characteristic of all adults who stutter, and it is certainly not a trait limited to those who stutter. To the extent it is a component of stuttering for many individuals, it probably interacts with a basic predisposition for difficulty with speech motor control.
To close out this review of emotional temperament in stuttering, I would like to call attention to a recent theoretical paper that touches on this topic. Walden, Frankel, Buhr, Johnson, and Conture (2012) make a strong case for the relevance of emotional factors in at least some children who stutter. They argue that stuttering may emerge as the result of two possible predispositions (a) emotional vulnerability and/or (b) limitations on rapid speech-language processing. Stuttering’s initial appearance and its day-to-day and moment-to-moment variability are a result of particular stressors or triggers for each of these predispositions. Triggers for the emotional predisposition involve environmental factors such as unexpected changes in the child’s day-to-day life or events that cause strong positive or negative emotion. The researchers support their position with a wide array of studies, many of which they have carried out, that demonstrate the existence of both predispositions and the effect on a child’s fluency of environmental stressors on each. Further evidence of the importance of environmental factors will be provided in Chapters 4 and 5.
Developmental, Environmental, and Learning Factors in Stuttering

Developmental Factors
Physical and Motor Skill Development
Speech and Language Development
The Onset of Stuttering
Delayed and Deviant Speech and Language Development
Cognitive Development
Cognitive Development and the Onset and Fluctuation of Stuttering
Cognitive Development and Reactions to Stuttering
Social and Emotional Development
Interference with Speech by Emotion
Stages of Social and Emotional Development
Emotional Security
Self-Consciousness and Sensitivity
Environmental Factors
Speech and Language Environment
Life Events
Learning Factors
Classical Conditioning
Classical Conditioning and Stuttering
Illustration of Classical Conditioning and Stuttering
Spread of Conditioning
Individual Differences in Conditioning
Undoing the Classical Conditioning
Operant Conditioning
Avoidance Conditioning
After studying this chapter, readers should be able to:
• Describe the analogy made between a computer with limited resources and achild learning to speak fluently
• Explain how factors related to physical development can interfere with fluentspeech
• Explain how factors related to speech and language development can interferewith fluent speech
• Explain how factors related to cognitive development can interfere with fluentspeech
• Explain how factors related to social-emotional development can interfere withfluent speech
• Describe the evidence for the role of parents, the speech-language environment,and life events in the onset and development of stuttering
• Describe how classical, operant, and avoidance learning may play a role in thedevelopment of stuttering
• Describe the role of learning new behaviors and unlearning old ones in thetreatment of stuttering
Competition for neural resources: The concept that the brain has a limited amount of resources that can be applied to tasks such as learning to speak and learning to walk. If some task requires a great deal of attention or uses a great deal of neural activity, other tasks performed at the same time will have fewer resources and may thus be less well performed
Speech and language environment: The communication style that characterizes people in a child’s environment–usually his home. For example, some parents, siblings, and other relatives of a child may speak very rapidly, use advanced forms or language, or interrupt the child frequently. These aspects of the speech and language environment are thought to stress the child
Life events: Happenings in a child’s life that may stress the child, such as parents’ divorce or being hospitalized
Classical conditioning: Repeated pairing of a neutral stimulus (such as a person) with a stimulus (a humiliating long stutter) that elicits a response (such as fear), so that the neutral stimulus eventually elicits the response. This process explains the spread of stuttering to more and more situations and words. Stuttering treatment can break this link with procedures like “desensitization,” which pairs the old behavior that elicited fear (e.g., a long stutter) with a different response (e.g., the clinician’s positive interest in the client’s long
Operant conditioning: Following a behavior with a reward or punishment so that the behavior becomes more frequent (if rewarded) or less frequent (if punished). Operant conditioning explains why stuttering behaviors become more and more abnormal as the child stutters more. But operant conditioning can be a powerful treatment tool as well
Avoidance conditioning: This type of learning occurs when a person uses a behavior to try to prevent an unpleasant occurrence by doing something; it is perpetuated by the successful prevention of the unpleasant experience, at least some of the time. In stuttering, avoidance conditioning may begin when a person first escapes from a stutter by saying an extra sound or word (like “uh”). Then he may make that sound even before saying the feared word, like “uh, can I have some pizza?”
Many factors, both in children themselves and in their outside environments, create the conditions under which stuttering first emerges and then either disappears or grows worse. Some of these factors are part of the child’s normal development, such as the explosive growth of speech and language skills during preschool years. Other factors may be common environmental situations that most children take in stride as they grow up, such as competing with siblings for attention and speaking time in a busy home. These developmental and environmental influences are not the cause of stuttering by themselves. Rather, they have their effects by interacting with the sensory-motor, language, and emotional factors discussed in Chapters 2 and 3 to precipitate stuttering sometime between the child’s acquisition of short phrases and puberty. Figure 4.1 depicts the influence of developmental and environmental factors (as well as learning) on the onset and development of stuttering.

Figure 4.1 Predisposing constitutional factors interact with developmental, environmental, and learning factors to precipitate or worsen stuttering.
The interaction of sensory-motor, language, and emotional factors with a child’s development and environment is ongoing, day by day. Such factors may have a gradual, cumulative effect, just like the forces of weather have on the surface of the earth. For many children, there is not a sudden landslide of stuttering, but a more gradual erosion of fluency. For others, stuttering appears out of a clear blue sky, then diminishes and may or may not reappear. Conditions at the onset of stuttering are typically not dramatic; the child is usually not under great pressure, nor has he just experienced a traumatic event. In my experience, stuttering often starts during ordinary stress–one or both parents away on a trip, relatives visiting for a holiday, a new baby in the house, or even no stress at all. The ordinariness of situations when stuttering first appears is reflected in this observation by Van Riper (1973, p. 81):
In the great majority of children we have carefully studied soon after onset, we were unable to state with any certainty … what precipitated the stuttering. In most instances there simply were no apparent conflicts, no illnesses, no opportunity to imitate, no shocks or frightening experiences. Stuttering seemed to begin under quite normal conditions of living and communicating.
Because children’s lives are often so normal when stuttering first emerges, research to determine critical developmental and environmental factors affecting its onset and progression has not produced notable results. This is a domain of educated guesses and tentative conclusions. Evidence for developmental factors is inferred from the fact that almost all onsets of stuttering occur when children are developing most rapidly during their preschool years (Andrews et al., 1983; Wingate, 1983). Evidence of environmental influences comes in part from clinical reports of particular stresses sometimes associated with the onset of stuttering and its remission when these stresses are lessened (e.g., Van Riper, 1973, 1982).Environmental factors are also implicated by higher incidences of stuttering in those cultures that are more competitive, with high standards and less tolerance of differences (Bloodstein & Ratner, 2008). Finally, some sources of evidence for genetic factors in stuttering are also evidence for environmental factors. These studies show that genes alone cannot account for the occurrence of stuttering in all children but rather some other factors–probably environmental–must also be responsible (e.g., Andrews, Morris-Yates, Howie, & Martin, 1991). However, this research has not been able to identify which environmental factors might be involved. The paucity of objective data tying specific developmental and environmental factors to stuttering will make this chapter more speculative than the last. Wherever I can, however, I will try to support speculations with facts.
In addition to developmental and environmental factors that may precipitate stuttering, there are influences that act upon stuttering once it begins. These are various types of learning–classical, operant, and avoidance conditioning. Learning, of course, affects all of our behaviors throughout our lifetime, and it may cause a child who initially stuttered only when he was excited, tired, or stressed to eventually stutter in ordinary circumstances, day after day. Learning can escalate mild, repetitive stuttering to severe blocking with a complex pattern of extra sounds, substituted words, and avoidance of speaking situations. Understanding the principles of learning will help you understand what may have created the individual stuttering patterns of clients. More importantly, this knowledge will guide treatment procedures as you help your clients acquire more effective communication behaviors and skills.
In this chapter, I have divided developmental, environmental, and learning factors into separate sections, although they do not operate independently; their actions are influenced by each other and by sensory-motor, language, and emotional factors, such as those described in Chapters 2 and 3. A developmental variable, such as cognitive level, may determine at what age avoidance conditioning may occur. An environmental variable, such as the stress of moving to a new neighborhood, may have different effects on different children, depending on such factors as a child’s sensitivity, physical development of speech production, and language maturation. It’s also critical to keep in mind when evaluating and treating clients that every client is an individual whose stuttering has evolved from unique contributions of developmental, environmental, learning, and constitutional factors.
Our view of how developmental factors affect children’s fluency assumes that there is in the growing child a competition for neural resources; that is, the brain must share its resources in coping with many demands. Like a computer, the brain can work on several things at once, but the more tasks it performs simultaneously, the more slowly and less efficiently it does each one. Unlike a computer, if the tasks are dissimilar, such as driving a car and talking about the weather, there is less interference between them. On the other hand, if the tasks are similar, such as rubbing your stomach while patting your head, there is more interference between them (Kinsbourne & Hicks, 1978). The problem of shared resources is more acute in children because their immature nervous systems have less processing capacity to share (Hiscock & Kinsbourne, 1977, 1980). Some children are especially at risk for straining their developing resources. Their speech and language skills may be delayed, yet they have to compete in a highly verbal environment. Or, their language development may surge ahead of their speech motor control skills, giving them much to say, but limited capacity to say it at a rapid rate, as though they were trying to push a crowd of people through a small door in a hurry. Such children may become excessively disfluent as other developmental demands outpace their abilities to coordinate the complex movements of rapid, articulate speech.
Here is an example of this competition between burgeoning language and slower motor abilities. Several years ago, I evaluated a 4-year-old girl whose uncle and grandmother stuttered. Her parents were concerned because she had been repeating words and sounds excessively for a year and a half, sometimes up to 20 times per instance. However, her language development was well above average; she began to talk with single words at 9 months and to produce sentences intelligibly at 12 months. In contrast, her motor development was somewhat slower; she had not walked until 18 months. I think it is possible that her disfluencies emerged as a result of a high proportion of her cerebral resources being used to formulate and express language with less mature capacities for motor activities, including fluent speech. In other words, a disparity between language facility and motor speech ability may have been an important contributor to the emergence of stuttering.
To appreciate how many skills and abilities the child is developing at the same time, look carefully at Figure 4.2. This chart covers only social, motor, and language domains, but it is clear that children have to master many different abilities simultaneously. If a child’s development is slower in one or more areas, her road to maturity may be steep and difficult at times. So, let us look at some of the domains of development and how they might contribute to the onset of stuttering.

Figure 4.2 Child development in the first five years. (Courtesy Harold Ireton, Ph.D.)
Physical and Motor Skill Development
The mother of a 3-year-old child who recently began to stutter told me “Whenever he has a spurt of growth, his stuttering seems to increase.” Why would this be? Between ages 1 and 6 years, children grow by leaps and bounds. Their bodies get bigger. Their nervous systems form new pathways and new connections. Their perceptual and motor skills improve with maturation and practice. This intensive period of growth is a twoedged sword for children predisposed to stuttering. Neurological maturation may provide more “functional cerebral space” that supports fluency, but it also spurs development of other motor behaviors that may compete with fluency for available neuronal resources. An example of such competition is the common observation that children learn to walk first or talk first, but not both at the same time. For example, Netsell (1981, p. 25) said of this trade-off, “The practice of walking or talking seems sufficient to ‘tie up’ all the available sensorimotor circuitry because the toddler seldom, if ever, undertakes both activities at once.” Likewise, Berk in his text on child development (1991, p. 194) suggested that “when infants forge ahead in spoken language, they seem to temporarily postpone mastery of new motor skills or vice versa.” Studies are needed to explore the specific effects on speech and language of mastering other skills. In Chapter 6, on theoretical perspectives on stuttering, I will introduce the “capacities and demands” view of stuttering, which may be a useful framework for understanding the competition for resources that I have been discussing.
The learning of motor control of speech by itself, even without acquisition of other motor skills at the same time, puts enormous demands on the child’s brain. Think about how difficult it must be to learn how to move the tongue, lips, and jaw to produce a desired sound when the relative size of these articulators is changing every day. Think about how the use of speech must draw on language resources as well. As Kent and Vorperian (2007, p. 73) remind us, “Speech development in children draws on a number of anatomical, motor, sensory, and cognitive resources. Ultimately, these various factors need to be integrated to account for a child’s progress toward the faculty of speech.” The integration of these factors may be a particular challenge for some children. Details about this process and evidence that individuals who stutter may be at a disadvantage are presented Chapter 5.
Before leaving the topic of motor skill learning, I would like to mention that another challenge to some children who stutter may be a delay in motor development. A review of studies of motor coordination in people who stutter can be found in Bloodstein and Ratner (2008). They cite several studies that have found children who stutter to be somewhat delayed as a group compared to nonstuttering children, but suggest it is not a clear-cut issue. I suspect that a significant delay in development of fine motor speech skills in a child with a strong urge to communicate and rapidly developing language abilities may set the stage for more serious disfluency. Such a delay may account for reduced speech intelligibility in some children, disfluency in others, and both problems in still other children. Given the evidence that boys are slower than girls in neuromotor development (Smith & Zelaznik, 2004), I suspect that boys are more at risk for the combination of reduced intelligibility and stuttering.
Speech and Language Development
The Onset of Stuttering
Although I discussed the interaction of speech-language development and the onset of stuttering in Chapters 2 and 3, it’s worth revisiting this topic in the context of overall development. Most stuttering begins between ages 2 and 4, a time when children acquire new sounds and learn new words almost by the hour (Bloodstein & Ratner, 2008). Yairi and Ambrose (2005) point out that many authors suggest that when stuttering begins before age 3, its onset coincides with “qualitative and quantitative advancements in the child’s articulation, phonology, morphology, and syntax” (p. 47). As with other developmental factors, we don’t believe that the demands of speech and language development cause stuttering, but that for many children, these demands may be a precipitating factor.
What is happening to precipitate stuttering during this speech and language growth spurt? You will remember from Chapters 2 and 3 that brain imaging studies of adults who stutter have shown some unexpected patterns of activity during speech. For example, there are abnormally high levels of activity in some regions of the right hemisphere and abnormally low levels in some areas of the left. In addition, recent findings with children who stutter suggest that areas of the brain used for integration of articulator planning, sensory feedback, and motor execution are compromised (e.g., Chang, Erickson, Ambrose, Hasegawa-Johnson, & Ludlow, 2008). Thus, planning and production of speech and language may use atypical neural pathways that may be slow or inefficient. But the demands grow ever greater as a child produces longer, faster, and more complex sentences. Tasks using different neural networks for segment selection, grammatical formulation, and prosodic planning must be orchestrated precisely so that each element is in place at the proper time as utterances are produced. If some components are ready but others are delayed, initial sounds or syllables may be repeated, prolonged, or even blocked, waiting for the whole sentence to be put together in the brain. A child’s reaction to this speech traffic jam may determine whether maladaptive learning occurs, or instead such moments soon disappear as the child develops compensatory strategies to pave the way to normal speech. We’ll look more at these reactions later.
Delayed and Deviant Speech and Language Development
Because it is thought that stuttering arises from constitutional differences (e.g., inheritance or congenital injury) in some children, it is natural to wonder if other speech and language abilities besides fluency are affected. It is also important to understand how delays in speech and language development might be related to the appearance of stuttering or disfluency. In general, research has found that speech and language delays or difficulties are more common among children who stutter than those who don’t, but the findings are neither simple nor clear-cut, and their implications are unclear.
More evidence of a possible relationship between language delays and stuttering is given in Chapter 5. Briefly, there are several possible relationships that can be hypothesized. One suggestion is that delays in language development may be related to stuttering because children with delayed language may become frustrated at their difficulty speaking, develop fears related to speaking, and thus learn to stutter as an anticipatory avoidance response (Bloodstein & Ratner, 2008). An example would be a child who has a gap between vocabulary development and growth of syntactic skills. This child may have the language form ready but unable to fill it with verbal content, making repetitions and even blocks possible as he tries to go ahead with speaking but is unable to. This may also happen to children who are not delayed, but advanced in one language component and typically developing in another.
Cognitive Development
I use the phrase “cognitive development” to refer to the growth of perception, attention, working memory, and executive functions that play roles in spoken language but are separate from it.
There may be two ways in which cognitive development affects stuttering. First, spurts in cognitive development may accompany the onset of stuttering as well as sudden increases in stuttering. Second, as a child who stutters develops more advanced cognitive abilities, he is more likely to become aware and even self-conscious of his stuttering.
Cognitive Development and the Onset and Fluctuation of Stuttering
Parents frequently report that the onset of their child’s stuttering occurred under the most normal of circumstances–no extra stresses in the household and no apparent increases in the child’s anxiety. The same “normal circumstances” are also frequently true for sudden changes in the ongoing stuttering of a preschool child–suddenly his stuttering is worse, and just as suddenly, it’s better. One factor that may not be obvious to the parent but nonetheless may be an influence on stuttering is the child’s cognitive growth. Earlier in this chapter, I suggested that aspects of physical development may affect stuttering; for example, learning to walk may make great demands on sensorymotor abilities, making fewer resources available for fluency. Now I am proposing that learning to think may make great demands on cognitive-linguistic abilities, leaving fewer resources available for rapid production of fluently spoken language.
This argument has been made before. Lindsay (1989) pointed out that during Jean Piaget’s “preoperational period” of childhood development from 2 to 6 years, a child goes through a series of transitions in which new cognitive learning must be assimilated and consolidated with current knowledge. These transitions are times when a child’s linguistic and cognitive systems are temporarily unstable before new concepts are mastered. As a consequence, children’s speech and language production during this period of adjustment may be vulnerable to disfluencies.
A number of authors have made other arguments that support a connection between cognition and stuttering, including the high frequency of disfluencies in children with developmental delays and traumatic brain injuries. These studies are summarized in Chapter 5.
Cognitive Development and Reactions to Stuttering
In the preceding section, I have suggested that children’s cognitive development may influence the onset of stuttering through competition for resources in the child’s brain. Now, I would like to argue that the role of cognitive development is also important in explaining how and when a child begins to form negative attitudes and beliefs about herself and her speech. Between ages 3 and 4, children’s cognitions mature enough so that they internalize the standards of behavior of those around them, including peers (Fagan, 2000). It is only at this point, according to Lewis (2000), that all children can evaluate how they are performing in comparison to others and will experience the “self-conscious” emotions of embarrassment, pride, shame, and guilt.
In regard to stuttering, once children who stutter compare their speech with others, they are likely to conclude that they are doing something wrong. Because some of the conclusions that children who stutter draw about their speech may come from other children’s reactions to their speech, a study by Ezrati-Vanacour, Platzky, and Yairi (2001) is of importance. These researchers looked at awareness of stuttering in typically developing children and found that some children were aware of stuttering in puppets at age 3, but most were not aware until age 5. Notably, most children at age 4 showed a preference for fluent speech, suggesting a negative evaluation of disfluent speech. Thus, peers of children who stutter may respond negatively to the speech of stuttering children at this age.
Emotions such as embarrassment and shame that arise from the increasing cognitive maturity of children who stutter and their peers may play an important role in the discomfort children feel when they stutter, and thus, it may affect the persistence of the stuttering. The embarrassment and shame some children feel about their stuttering are probably the most important cognitive-emotional factors that give rise to increases in tension, escape, and avoidance responses that make stuttering a self-sustaining disorder and increasingly difficult to overcome. In my experience, most children who recover completely from stuttering with treatment are younger than 5, perhaps a significant age, given the evidence cited above about peer awareness of stuttering.
Social and Emotional Development
Interference With Speech by Emotion
At some time in your life, you have probably experienced the effects of strong emotion on your speech. If you’ve been nervous when talking in front of an audience, your voice may have quavered or you may have been talking faster than you meant to but couldn’t help it. When you get really worked up, like when you have to make a phone call in an emergency, rapid breathing and tension in your larynx can make it difficult to talk. The same sort of interference by emotion may be even more prevalent in early childhood, because a child’s speech and language neural networks and structures are immature, not fully myelinated, and may not be buffered from “cross talk,” or interference by the limbic (emotional) system structures and pathways involved in the regulation and expression of emotion. Such interference may be even more likely among many children predisposed to stutter. Their slower maturing speech and language functions may not be optimally localized or adequately insulated from interference and may be closer to centers of emotion in the right hemisphere, a hypothesis I discussed in Chapter 2. Thus, when such children are emotionally aroused, fluency may suffer because neural signals for properly timed and sequenced muscle contractions may be interrupted in some way. I see evidence of this when I ask parents when their child first began to stutter. They frequently tell me that they noticed stuttering for the first time when their child was highly excited about something.
Excitement is commonly mentioned in the literature as a stimulus that elicits disfluency. Starkweather (1987) noted that “all children speak more disfluently during periods of excitement.” Dorothy Davis (1940), who conducted one of the first studies of normal disfluency, reported that of the 10 situations in which children showed repetitions in their speech, “excitement over own activity” was when they most frequently repeated sounds and words. In a later study, Johnson and associates (1959) asked parents of children identified as stutterers to describe the situation in which they first observed their child’s stuttering. They most often reported that the first appearance of stuttering occurred when the child was in a hurry to tell something or was in an excited state. Thus, both stuttering and normal disfluency seem to occur most often or noticeably during states of transitory emotional arousal.
Stages of Social and Emotional Development
Some stages of development may provide more social and emotional stress than others. For example, the processes of separation and individuation are known to be periods of stress. After a child passes his second birthday, he strives harder for autonomy, creating the conflicts of the “terrible twos.” Most parents gradually relinquish control, while at the same time helping their child learn the limits of his freedom. In some cases, however, the transition from a dependent infant to an independent preschooler may occur too rapidly for a parent or a child. If the child is pushed toward independence faster than he wants, he may feel frustrated and insecure because his mother seems less nurturing. A mother may become alarmed if she isn’t ready for her child’s quest for independence and may try to restrain her child. A child may conform to these restraints but feel angry and frustrated. Yet, the child cannot easily express these feelings to someone he depends on so much, and disfluency may result in those interactions in which such emotional ambivalence and conflict affect motor control of speech (Lidz, 1968).
Emotional Security
As a young child grows older, other members of the family besides the mother play a role in social and emotional changes. Although a child’s father and siblings comprise a wider support system, a child’s resentment at having to share his mother’s attention may elicit feelings of anger, aggression, and guilt. It seems possible that if such feelings are punished or ignored, transient disfluency or more severe stuttering may result in some children.
One of the more common provocations for feeling resentment is the birth of a sibling. I discuss the effect of a sibling’s birth on fluency later in the section on environmental factors, but it warrants mentioning here too because a child’s strong emotions may often reflect his developmental level as well as the environmental event that triggered the emotions. Theodore Lidz (1968, p. 246), a developmental psychiatrist with interests in speech and language, provided this example:
Psychoanalytically oriented play therapy with children also indicates that many of their forbidden wishes and ideas have relatively simple access to consciousness. A 6-year-old boy who started to stammer severely after a baby sister was born was watched playing with a family of dolls. He placed a baby doll in a crib next to the parent dolls’ bed and then had a boy doll come and throw the baby to the floor, beat it, and throw it into a corner. He then put the boy doll into the crib. In a subsequent session, he had the father doll pummel the mother doll’s abdomen, saying, “No, no!” At this point of childhood, even though certain unacceptable ideas cannot be talked about, they are still not definitely repressed.
Although we feel, as does Lidz, that stuttering may be triggered by the birth of a sibling, our belief about the underlying cause is not so Freudian. Many threats to feelings of security can create emotional stress that may disrupt the speech of children who are predisposed to stutter. As we will see in the section on treatment, we have found that therapy strategies that increase a child’s sense of security and help him learn to speak more fluently will suffice for many children who begin stuttering under these emotional stresses.
Self-Consciousness and Sensitivity
As we noted earlier in the section on cognitive development, the emergence of selfconsciousness, which begins during the child’s second year, may be another source of social and emotional stress. This reflects the child’s growing awareness of how he is performing relative to adult expectations. Although this process is not thoroughly understood, Jerome Kagan presents an interesting description of it in his book, The Second Year (1981). In a relevant example, Kagan proposes that the self-corrections a child makes in his speech are evidence of this self-awareness. Taking this further, we can surmise that increasing self-awareness in a child who is excessively disfluent might lead to self-corrections and stoppages that only worsen the problem.
In Chapter 2, we briefly discussed the hypothesis that people who stutter, as a group, may have unusually sensitive temperaments. Research on temperament in nonstuttering children, especially the longitudinal studies of Calkins and Fox (1994) and Kagan and Snidman (1991), indicates that the social-emotional traits of fearfulness and withdrawal that accompany more sensitive temperaments can change over the course of a child’s preschool years. Some children become better able to modulate their temperamental tendencies, but others remain hostages to their temperaments. Such individual adaptations may be crucial in determining which children who begin to stutter will continue to do so and which will stop.
Recent reports on sensitive temperament have included the positive as well as negative aspects of this personality type. In an overview article, Ellis and Boyce (2008) suggest that a reactive temperament in a child may produce very different outcomes, depending on whether a child encounters a stressful environment or a nurturing one. For example, Boyce, Chesney, Alkon-Leonard, Tschann, Adams, and colleagues (1995) found that, compared to relatively unreactive children, sensitive children in stressful environments had more respiratory illnesses than typical children, whereas sensitive children in nurturing environments had fewer. In other words, a sensitive temperament can give a child protection against illness if the environment is favorable. What are the implications for stuttering? Future research may show that sensitive children with family histories of stuttering who experience stressful childhoods develop persistent stuttering. But those with the same background who grow up in highly supportive environments recover from any stuttering they have in early preschool years. And those with family histories of stuttering but an unreactive temperament may have a mix of persistence and recovery. This view of children who stutter who also have sensitive temperaments suggests that treatments may be most effective if they can help parents create maximally supportive, nurturing environments.
Before leaving this section, I want to comment on stutterers’ psychological adjustments in general. Many people who have little exposure to stuttering believe that stutterers are essentially nervous people or that stuttering is a sign of neurosis. If this were true, we should have found evidence of psychological maladjustment or excessive anxiety in people who stutter, particularly when stuttering first begins in childhood. However, research on the personalities and adjustments of people who stutter has found no convincing evidence that they differ from nonstutterers in these ways (Bloch & Goodstein, 1971; Bloodstein & Ratner, 2008; Van Riper, 1982). A few findings suggest that adults who stutter may not be as socially well adjusted as those who do not, but this may reflect the influence of stuttering on social experiences (Bloodstein, 1987; Bloodstein & Ratner, 2008).
Summarizing the effect of social and emotional development on fluency, I have suggested that many of the normal social and emotional stresses that children experience may result in disfluent speech, although the evidence is mostly anecdotal. Moreover, we suspect that children who are neurophysiologically vulnerable to stuttering may be especially prone to difficulty when social conflicts and emotions create extra “noise” in their neural circuitry for speech. This would be particularly true for those children who are both predisposed to stuttering and who have emotionally reactive temperaments. In general, however, children who stutter appear to be as psychologically well adjusted as those who do not stutter, despite the extreme emotional stress that stuttering itself can impose.
Some children who have predispositions for stuttering may show initial signs of stuttering in response to developmental pressures alone, but most who become stutterers are probably affected by environmental pressures as well. Such pressures typically result from attitudes, behaviors, or events that occur in their homes. One example is a family’s anxiety about the child’s speech, which may be readily apparent in the facial expressions of parents and siblings when the child is disfluent. Another pressure may be the conversational style in the child’s home that is distinguished by lots of interruptions as well as rapid, complex speech that is beyond the child’s level. There are many things in a child’s environment–some subtle and some not so subtle –that can add enough stress to a child with a predisposition to stutter to trigger the onset and promote the growth of stuttering.
When I described the effects of developmental factors on a child’s speech, I used the analogy of a computer that is overloaded by too many simultaneous tasks. The computer analogy is also appropriate to depict environmental factors, but now you need to imagine that a computer being used for programs that exceed its capacities is also being subjected to periodic power surges or external commands that its programs cannot process. It is easy to imagine that similar circumstances would likely create fluency breakdowns in a vulnerable child. These kinds of environmental pressures can also worsen the symptoms of stuttering, as I describe in Chapter 5. I will begin this discussion of environmental factors by reviewing research on the most important factor in family environments, the parents.
Much of the early research on parents of children who stutter was done in the 1950s at the University of Iowa. These studies suggested that parents of children who stutter are more critical and perfectionistic. This perspective shaped the public’s view of stuttering for many generations and still has some influence. For example, parents of the children I work with sometimes tell me in our first meeting that they think they may have caused their child’s stuttering by being too demanding of him when he first started to speak.
Valid criticisms of this early research point out that these studies didn’t use control groups of parents of children with other disorders; thus, we can’t dismiss the possibility that rather than causing stuttering, these differences in parenting may be the result of stuttering. This possibility is all the more likely given that these parents were interviewed more than a year after the onset of their child’s stuttering. Perhaps any parent of a child with a disorder would appear to have high standards for him. Another objection to the claim that parents of children who stutter caused the stuttering by their perfectionism is that their brothers and sisters would have stuttered also, but this appears not to have been the case.
More recent studies of parents of stutterers and parents of nonstutterers–reviewed in Chapter 5–produced mixed results. Some indicate that parents of stutterers are more anxious or more rejecting, while others find no differences. On balance, it seems likely that some children who stutter grew up with parents who were a little more demanding or anxious than average, and this may have made a difference. Assuming they already had a constitutional predisposition to stutter, their early speech may have been peppered with disfluencies, which would have alarmed their parents. The children then, picking up their parents’ dismay, may have become self-conscious about their minor disfluencies and thus have unwittingly added tension and struggle, which blossomed into more noticeable stuttering.
It is interesting to speculate that in some cases a child’s hypersensitivity to parents’ concern and their increased tension as a response to their disfluencies is a component of an overall vulnerable temperament found in some children who stutter (Anderson, Pellowski, Conture, & Kelly, 2003; Karrass et al., 2006; Oyler & Ramig, 1995), which may be inherited. Their parents, the genetic source of these temperaments, may be the anxious, overprotective mothers or fathers that have been described in the literature on parents of stutterers. Such children are, therefore, in double jeopardy for persistent stuttering because of their own temperaments and because one or both parents are overly concerned about their children’s stuttering because of their own temperaments. On the other hand, some parents may have an ameliorating effect on a child’s vulnerable temperament, making it possible for a child who begins to stutter and who is emotionally reactive to recover from stuttering. In a discussion of environmental influences on biological predispositions of children who do not stutter, Calkins and Fox (1994, p. 209) said that “the child’s interactions with a parent provide the context for learning skills and strategies for managing emotional reactivity.”
Once again we see that influences on stuttering are numerous and complex, coming from both the child and the environment. Some of these influences may precipitate stuttering, others may interact to make remission difficult, and still others may provide the kinds of support that make remission possible.
Speech and Language Environment
Because every preschool-age child is tuned into the speech and language around him, especially that of his parents, the communication style surrounding him may be an important influence on the child who stutters. Writers have hypothesized that as a child tries to emulate adult models of speech and language, to use longer words and longer sentences, to try less familiar words, and to pack more meaning into his utterances, he will be more likely to stutter. Van Riper (1973) expressed it this way: “Stuttering usually begins at the very time that great advances in sentence construction occur, and it seems tenable that, when the speech models provided by the parents or siblings of the child are too difficult for him to follow, some faltering will ensue” (p. 381). Later, in the same volume, Van Riper goes on to cite nine references in which clinicians point to parental speech models as a major source of stress on a child’s fluency (pp. 380–383). This stress includes not only the parents’ speech and language but also the conditions under which the child tries to speak.
Following in Van Riper’s footsteps, many other clinicians have speculated that speech and language environments are a potential source of stress for children who stutter (e.g., Gottwald, 2010; Richels & Conture, 2007; Shapiro, 1999; Starkweather, Gottwald, & Halfond, 1990; Zebrowski & Kelly, 2002). Table 4.1 lists a variety of sources of possible communicative stress.
Table 4.1 Possible Speech and Language Stresses

Two writers have moved beyond clinical speculation to develop informal theories about the influence of adult models on a child’s fluency. Crystal (1987) proposed an “interactive” view of many speech and language disorders, which suggested that demands at one level of language production (e.g., syntax) may deplete resources for other levels (e.g., prosody or phonology) and result in breakdown. His supporting data nicely illustrate how stuttering may be exacerbated by a child’s use of advanced language. He presented evidence that the more complex the syntax and semantics that a child used, the more he stuttered. Starkweather (1987), describing a demands-andcapacities view of stuttering, commented that “the production of speech and the formulation of language place a simultaneous demand on the young person. If the demands in either of these two dimensions are excessive, performance in the other dimension may be reduced.” These two views imply that stuttering may increase when an individual uses longer words, less frequently occurring words, more informationbearing words, and longer sentences. Stuttering may also increase when the individual is uttering a more linguistically complex sentence. By implication, the child’s speech and language environment–usually conversation by adults talking to the child– may be responsible for influencing a child to use more advanced language.
What do we know about the speech and language of parents of children who stutter? Research on this topic is summarized in the appendix, and all citations are given there. It is important to note that not all of the conclusions given below are supported by every study, but there are some generalizations we can make about four aspects of parents’ speech that are thought to increase stuttering: speech rate, frequency of interruptions, frequency of questions, and complexity of language. Specifically, (1) speech rates of parents of children who stutter may be faster, and this may be more reliably so for parents of children who are severe stutterers. (2) Mothers of stutterers and even mothers of nonstutterers who interacted with children who stuttered interrupted the stuttering children more frequently when these children were stuttering than when they were fluent; additionally, the durations of the mothers’ interruptions may be related to the severity of the child’s stuttering. (3) Parents of children who stutter may ask about the same number of questions as parents of nonstuttering children, but when children give longer answers to questions, they stutter more. (4) For children who began to stutter, the more complex the mothers’ language, the less likely the children are to recover naturally.
What are the clinical implications of this view? Even the most cautious researchers examining the parent-child interactions of stuttering children are hopeful about the possible therapeutic value of changing some parents’ verbal behaviors. They advocate clinical research on the therapeutic effects of reducing parents’ speech rates (Nippold & Rudzinski, 1995), changing their language patterns (Miles & Ratner, 2001), or even combining these approaches with direct treatment of children’s fluency.
Life Events
Certain life events can deliver a blow to a child’s stability and security. When this happens, stuttering may suddenly appear out of nowhere, or previously easy repetitions may be transformed into hard, struggled blocks. To have someone close to you die, to be hospitalized for an operation, or to have parents divorce is difficult for any of us, but it is especially difficult for children. Obviously, many children go through such events and adapt to them without apparent major problems. But children who are predisposed to stuttering often show the effects of such events in their speech. Kagan (1994a) noted that some children who begin life with relaxed temperaments may even become shy and fearful under the onslaught of stressful events. This may well set the stage for stuttering if other constitutional factors predispose the child for it. You may wonder what the mechanism is by which stress can precipitate or worsen stuttering. I don’t know the answer, but it seems likely that if a child’s brain pathways for speech and language are slightly compromised, extra resources are needed to maintain fluency. When stress increases negative emotions such as anxiety, it seems possible that the negative emotions would consume the extra resources. However, it still leaves us with an incomplete picture. How exactly do extra resources compensate for deficits? How does anxiety consume extra resources?
There is little research on the relationship between stressful life events and stuttering, but many authors have observed the connection. Starkweather (1987), for example, wrote, “All children speak more disfluently during periods of tension–when moving or changing schools, when their parents divorce, or after the death of a family member” (pp. 146–147). These increases in disfluency could easily result in the onset of stuttering or in increased stuttering in children who are vulnerable to such stresses. Johnson and associates (1959) noted that the following events were among the 16 situations in which parents first noticed their child’s stuttering: (a) child’s physical environment changed (e.g., moving to new house); (b) child became ill; (c) child realized his mother was pregnant; (d) a new baby arrived. In discussing the onset of stuttering, Van Riper (1982) acknowledged that various studies have found no differences in the amount of emotional conflict in the homes of children who developed stuttering versus those who didn’t. However, he went on to note, “Nevertheless, we have studied individual cases in which stuttering did seem [to be] triggered by such conflicts, and it is difficult for us to ignore these experiences” (p. 79).
My own clinical experience is similar. In the past several years, for example, during which I’ve evaluated dozens of children who stutter, I’ve encountered four children in four different families who began to stutter when their parents were in the early stages of divorce. However, this turmoil was not the only factor in their stuttering. Three of the children had relatives who stuttered, and the father of the fourth child was a stutterer. Moreover, all four were preschoolers and were probably experiencing various growth and development pressures. But for all of them, their parents’ divorce appeared to be a factor that pushed them from normal speech to stuttering.
In another case of a life event–precipitating stuttering, I evaluated a 9-year-old girl who began to stutter when her classroom teacher had an emotional breakdown that was apparent in the classroom. The teacher’s outbursts of anger and crying, interspersed with high demands for rapid performance on frequent examinations, were apparently extremely stressful for this student. Under this stress she developed tight blocks, with physical tension at the level of the larynx and abdomen. Even though I was convinced through extensive interviews with the family that the child had no prior stuttering, I noted several predisposing factors for stuttering. First, her younger sister had significant learning disabilities, including auditory processing problems. Second, her mother described herself and her daughter who stuttered as shy and emotionally reactive. These two factors–a family history of learning disability and a vulnerable temperament–may have provided a fertile matrix for the sudden germination of stuttering when a stressful life event occurred. Happily, after a year of treatment, this child became fluent.
In a few cases, traumatic life events appear to precipitate stuttering in children as well as in adults who appear to have no predisposition to stuttering. These unusual onsets are discussed in Chapter 15 when we explore “psychogenic stuttering.” Such individuals often stutter in unusual ways, which differ from the “garden variety” stuttering of those who begin stuttering when a stressful life event interacts with various predisposing conditions. Table 4.2 lists some of the life events that we have found to be stressful to children’s fluency.
Table 4.2 Stressful Life Events That May Increase a Child’s Disfluency

When I was in high school, I stuttered severely. If I had to make a phone call, I dreaded it for hours beforehand. When I finally got up the courage to pick up the phone and dial, an invisible hand seemed to tighten around my throat. When someone answered, all I could get out was a series of “ums” punctuated by the listener’s repeated “Hello? Hello? Hello?” Finally I could say “Is Mmmmmm” and there I was, stuck fast, until with a huge head jerk I would blurt out, “Is Molly there?” And the listener would say, typically, “What?”
I had gone downhill a long way since my stuttering began at age 3 with simple repetitions of words like “I-I-I-I…” So, how did this change take place? The answer is learning–the change that takes place in a person or animal as a result of their experiences in the environment (Lefton, 1997). Learning created my more severe stuttering symptoms as I went through elementary school, junior high, and high school. Learning also helped me–with the guidance of a good stuttering therapist–to reduce those symptoms to the mild form of stuttering that I have today. In this section, you will learn more about how learning works, so that you will understand your client’s stuttering behaviors and then can help him change them. The different components of my stuttering were created by different types of learning. Specific types of learning are often referred to as different kinds of conditioning. These include classical conditioning, operant conditioning, and avoidance conditioning.
Classical Conditioning
Classical conditioning was first scientifically described by the Russian physiologist, Ivan Pavlov. As with many discoveries, it was serendipitous. Pavlov was actually studying how dogs digested their food, which he did by measuring their saliva when they were fed. One morning he happened to notice that the dogs were salivating even before they were fed, seemingly in response to just seeing him walk in the door of his laboratory. This observation led him to abandon his study of dogs’ digestion and begin a series of studies of how the anticipatory salivation worked. Would any sight or sound occurring just before the dogs were fed elicit salivating? First, he rang a bell just before the dogs were fed; then, he rang the bell without food. At first the bell alone didn’t elicit anticipatory salivation. But then, like a good scientist, he persisted. His intuition about his earlier observation led him to carry out many pairings of the bell and subsequent feeding. At last he saw results; after many pairings (conditioning), the bell elicited salivation even though the food was nowhere in sight.
Pavlov’s observation provided the first scientific understanding of classical conditioning. Since then, classical conditioning has been studied extensively, and scientists have been able to describe how it takes place. Figure 4.3 depicts the “paradigm” (a model or diagram of how a process takes place) for classical conditioning.

Figure 4.3 Classical conditioning paradigm. Dog salivates naturally when given food. Food is paired frequently with sound of bell. Bell without food then elicits salivation.
For classical conditioning to take place, several things must occur:
• A stimulus that reliably elicits a response must be present. This stimulus is called theunconditioned stimulus (UCS), and the response it elicits–often a reflexive or hardwired response–is called an unconditioned response (UCR).
• Then a neutral stimulus that doesn’t elicit any particular response must be pairedwith the UCS. The neutral stimulus is called the conditioned stimulus (CS) because it will be conditioned to elicit a response.
• After repeated pairing of the CS with the UCS (which reliably elicits the UCR), theCS is then presented without the UCS, and voila! The CS elicits the UCR.
Classical Conditioning and Stuttering
Remember my description of my stuttering in high school? The feelings of dread I had before making a phone call and the tightening of my throat were the result of classical conditioning. The neutral stimulus (CS) was the phone or the thought of making a phone call. Because I had had lots of experience with stuttering at that time and it was all unpleasant, embarrassing, and shameful, the expectation of stuttering (UCS) elicited strong negative emotions for me (UCR). Those emotions, like many people’s fear of public speaking, resulted in considerable tension in my larynx and other muscles related to speech. Repeated experiences that paired the phone with the experiences of stuttering conditioned me to experience fear when contemplating making a phone call. The fear elicited muscle tension that undoubtedly worsened my stuttering every time I made a phone call. I think my experiences are typical of how classical conditioning can make stuttering worse for many people who stutter.
An excellent theoretical account of classical conditioning and stuttering was provided by Eugene Brutten, a speech-language pathologist, and Donald Shoemaker, a psychologist, who worked together at the University of Southern Illinois (Brutten & Shoemaker, 1967). They hypothesized that the earliest stuttering symptoms (often repetitions) result from the cognitive and motor disorganization that occurs when a child’s anxiety is conditioned to speech.
Although the description of classical conditioning and stuttering I am giving here owes much to Brutten and Shoemaker’s (1967) pioneering work, I believe classical conditioning is seldom responsible for the earliest signs of stuttering. Instead, I believe, along with Starkweather (1987, p. 372), that “it seems likely that [neuro]physiological sources play more of a role in stuttering onset, whereas conditioning processes play more of a role in stuttering development.” I also agree with a similar assessment by Van Riper (1982) that “the real contribution of classical conditioning theory as it is applied to stuttering lies in its ability to explain the development of the disorder.”
Illustration of Classical Conditioning and Stuttering
Let me give you a description of how classical conditioning might actually work in the development of childhood stuttering. For ease of explanation, I will refer to hypothesized increases in muscle tension (and perhaps speech rate) during disfluencies under conditions of negative emotion as “the tension response.” The paradigm to illustrate this conditioning is shown in Figure 4.4.

Figure 4.4 Example of classical conditioning of tension in a beginning stutterer. The repeated pairing of easy stuttering with a critical listener or internal frustration, which elicits tension, makes the easy stuttering a conditioned stimulus for tension. Consequently, tense stuttering occurs in more and more situations.
To illustrate how this paradigm applies to an individual child who stutters, I will use a hypothetical 4-year-old I’ll call Richard.
In this example, Richard’s stuttering up to this point has been relaxed, slow partword repetitions. This will be the CS, but it hasn’t been conditioned yet, so it’s still neutral. When he experiences this easy stuttering, Richard doesn’t respond to it in any particular way. It doesn’t bother him, and he doesn’t react to it. As I describe this process, I will sometimes refer to the CS as “easy stuttering” or as “disfluency.” The point I’m trying to make here is that a child’s early stuttering may often go unnoticed and does not include tension or hurry. But with conditioning, it becomes tense and hurried.
Now we observe Richard as he goes along in his daily life, growing up and becoming more aware of how he is performing compared to other kids. Different things may happen to pair the neutral stimulus (Richard’s easy stuttering) with a UCS (something that causes Richard to react to his stuttering in a way he didn’t previously). This pairing will turn the neutral stimulus into a CS. For example, Richard may notice that as he is stuttering, he is not getting his words out as fast as he’d like compared to his peers or his own internal expectations. This may cause an emotional response, such as frustration. With frustration comes an increase in his body tension, especially in speech muscles (see the section on temperament in Chapter 6 for a full description of this). This is Richard’s UCR. Other things may also occur–for example, Richard’s stuttering may elicit negative responses from his parents or playmates, causing him to feel embarrassed or ashamed. His father may frown and suggest impatiently that he stop speaking and start again. Richard’s emotional response to this may also trigger a tension response–tightening speech muscles and repeating sounds a little faster. The important thing to understand is that this repeated pairing of the previously neutral stimulus (stuttering) with the UCR (tension response) will turn the first moments of stuttering and even the anticipation of stuttering into a CS that elicits the UCR.
Other factors influencing Richard’s UCR may be in his external environment. For example, Richard may have a new baby brother or his family may have just moved to a new town. These factors may increase a child’s anxiety (autonomic arousal), particularly in a sensitive child, and may thus make the UCR (tension response) more likely and stronger and also make the learning more powerful. This will be described at greater length in the later section on individual differences.
Now, the experience of stuttering or the anticipation of stuttering elicits the tension response, even in the absence of any threatening stimuli. Previously, Richard used to stutter with only easy disfluencies in friendly situations like talking with his mother or playing with a peer; now he stutters with tension and hurry in those situations because his stuttering (CS) or anticipation of it triggers the tension response. This conditioning process also involves generalization of Richard’s stuttering to more and more situations, involving a process that we will describe in the following section.
Spread of Conditioning
Conditioning is an active and continuing process. When a child is disfluent (CS) and experiences the tension response (CR) (a CS elicits a conditioned response), a host of other stimuli are present, incidentally. When the child stutters, he is talking to someone, uttering a particular word or sound, speaking in a particular room, and talking about a particular topic. Because of the power of classical conditioning, the pairing of these other stimuli with the CS (disfluency) gives them the potency to elicit the CR. The particular sound on which he stutters, for example, may become conditioned to elicit the CR, so in the future he is more likely to stutter on that sound with tension and hurry. Thus, as conditioning takes place again and again, the stimulus becomes a complex of many things, including words and sounds, listeners, and physical surroundings or situations. This chaining of stimuli is called “higher-order conditioning” or “second-order conditioning.” Figure 4.5 illustrates the process.

Figure 4.5 Spread of conditioning. (1) Easy disfluency repeatedly elicits the tension response, producing tense stuttering. (2) Easy disfluency, which occurs with various words, situations, sounds, and people, is then a conditioned stimulus that elicits the tension response and tense stuttering. (3) Words, situations, sounds, and people become the conditioned stimuli, which elicit the tension response at the onset of an utterance, producing fixed articulatory postures (or blocks) when the child begins an utterance.
It is important to note that to maintain the effects of the CS, a UCS must occur periodically. Stuttering must occasionally provoke frustration or another negative emotion for it to continue to elicit the tension response.
The spread of conditioning to other conditioned stimuli results in changes in stuttering as well. Initially, a child might emit several repetitive disfluencies or a long prolongation before the tension response occurs. Soon, however, muscle tension occurs earlier and earlier in the stutters. When other conditioned stimuli, such as words, elicit the tension response, the easy repetitive disfluencies may not occur at all. Instead, a child will increase muscle tension on the very first sound he tries to utter, resulting in the “fixed articulatory postures” that are a sign of advancing stuttering.
As a child’s stuttering frequency increases as a result of the spread of conditioning to more and more stimuli, the duration of the child’s stuttering may also increase. This may be explained by the fact that the tension response soon becomes a stimulus that elicits more tension. After all, the tension response makes it harder to utter a word, and the experience of “squeezing hard” without being able to speak for a second or two elicits frustration, leading to another tension response.
Individual Differences in Conditioning
Before we leave our discussion of classical conditioning, I would like to touch on the topic of individual differences and conditioning. The rapid learning and widespread generalization that is apparent in stuttering may parallel the “prepared classical conditioning” of some animals that are rapidly and deeply conditioned to such naturally dangerous objects as snakes (Mineka, 1985). In humans, rapid conditioning can occur in those with temperaments that are more alert to threatening stimuli. For example, individuals with an anxious temperament are more prone to acquire fears and phobias (Biederman, Rosenbaum, Chaloff, & Kagan, 1995), and animals, which have been exposed to stressful situations, are more easily classically conditioned (Shors, Weiss, & Thompson, 1992). Thus, it seems likely that children who stutter and who are especially sensitive may rapidly condition to such threatening stimuli as a critical parent or a peer who makes fun of the child’s stuttering.
Undoing the Classical Conditioning
How can you use the principles of classical conditioning to help someone who stutters talk more fluently? Treatment of the conditioned tension response in young children can involve behavioral therapy that gives the child enough experience with fluent speech that the old conditioned response is finally extinguished. For this to occur, fluency-facilitating conditions are set up on a daily basis. This may involve a one-onone relaxed conversation with a parent, starting with simple language so that the child repeatedly has conditioned stimuli (parent, anticipation of speaking) followed by the absence of the conditioned tension response. Instead, the conditioned stimuli are associated with fluent speech. This repeated pairing is first done with one parent and then with other members of the family, as well as situations outside the home so that gradually, fluency is generalized to many situations. Operant or instrumental conditioning is an important component of this process and will be described in an upcoming section.
When the conditioned tension response has been deeply learned, as is often the case with older children, adolescents, and adults, considerable unlearning must take place. The clinician has a number of options, but one of several tools I use frequently is desensitization. Remember that when I described the original conditioning, I suggested that there is an emotional reaction, such as frustration, that triggers a UCR of tensing and speeding up. Desensitization aims to reduce the emotional response by helping the client go ahead and stutter but feel emotionally accepting of it and relaxed about it. This is achieved by rewarding the client for staying in the stutter and staying calm while in it, decreasing the emotional response. Once the client can do this in many situations, the client is helped to stutter in a new way that sounds like the typical speech of an average speaker. A full description of this aspect of treatment will be given in the chapters on treating school-age children (Chapter 13) and adults (Chapter 14).
Because the CS has been associated with many different people, situations, and words, undoing the conditioning must involve pairing many different conditioned stimuli (people, situations, words) with the new responses.
Operant Conditioning
When I stuttered severely in high school, I would often end stutters by jerking my head just as I made a big push that finally got the word out. Operant conditioning was responsible for my learning this behavior and for the fact that it remained part of my stuttering pattern for many years. In this type of learning, the frequency at which a behavior occurs is related to the consequences that follow. If a behavior is followed by a reward, it increases; if it is followed by an aversive consequence, it decreases. Like many operant behaviors, my head nod had begun as random struggles when I was jammed on a word. Several times the word I was stuck on popped out of my mouth just as I jerked my head up. The relief I felt in freeing myself of the stutter and being able to finish the word caused my head jerking to increase whenever I stuttered. The escape behaviors described in Chapter 1, such as eye blinks, are learned through operant conditioning. They begin as part of random struggle efforts to escape from a stutter, and they are rewarded by the release of the word.
A common type of operant conditioning is called positive reinforcement. The next time you’re in a public building with an elevator, watch what happens when people are waiting for the elevator to arrive and it’s delayed. One person will usually keep pressing the elevator button several times even though the button is illuminated and the elevator is on its way. Like everyone else, this person has been reinforced many times by the prompt arrival of the elevator after pushing the button just once. At some later time, however, he must have encountered a delayed elevator, and after pressing the button repeatedly, he was rewarded by the eventual arrival of the elevator. Thus, positive reinforcement conditioned him to repeat this behavior even though he had to have known that it wasn’t necessary. The same is true if you have a “lucky shirt” to wear to an exam. If you do well, you will probably be more likely to wear the same shirt–even if it is in need of a wash–to your next exam in spite of intellectually knowing that it doesn’t actually affect the outcome.
Another kind of operant conditioning is punishment. It has probably shaped your behavior if you’ve ever gotten a ticket for speeding. The punishing effect of an expensive fine makes you drive more cautiously when you are on the same section of road. You can appreciate the disadvantage of punishment, however, if you can remember referring to the police officer as a “weasel” when you described the incident to your friends and family. Unless it is mild and delivered with good humor, punishment usually is not beneficial to clinical relationships.
A third kind of operant conditioning, negative reinforcement, is the mechanism behind the escape behaviors, like eye blinks and head nods, that are part of many stuttering patterns. The head jerks that were part of my own stuttering pattern were the result of just this type of conditioning. Negative reinforcement occurs whenever a behavior is followed by the termination of an unpleasant situation. In stuttering, a sudden movement may terminate a moment of stuttering, which means that behavior is reinforced and is likely to occur more frequently in the future when the stutterer is jammed up. All of us may develop habits through negative reinforcement in our daily routines. For instance, if you are trapped in a long line at the cash register in the bookstore, you may bail out and discover that the register upstairs in the CD and DVD section has no line at all. If so, your escape behavior is reinforced, making it more likely that you will again go upstairs to the other register the next time you are stuck in the long line.
Operant conditioning is a major tool for all clinicians. It is a component of every stuttering treatment program. It’s a very powerful tool that affects the frequency of the behavior you are interested in, especially in the environment in which you use it. For example, when you are working with a preschool-age child who stutters, one effective approach is to help parents arrange a situation in which the child is fluent, then praise her to reinforce her fluency. After that has been done many times, the frequency of the child’s fluent utterances will increase. Remember that the environment in which the operant conditioning takes place is important. Therefore, to increase the child’s fluency where it counts most, the clinician teaches the parents to elicit fluency and praise it in their home. With a child, very mild punishment might be used occasionally to help her learn to use fluent speech instead of stuttering, but it must be done with great care and careful monitoring of the effect. Whenever punishment is used, reward is also an important part of a treatment plan. Much reward and little punishment is a good motto. Van Riper (1973) pioneered the use of mild self-punishment to be used by adults who stutter, using a procedure he called “cancellation.” Van Riper taught clients to stop talking immediately after they had one of their typical tense stutters and then in silence, plan an easier, more relaxed way of stuttering on the word. That silent pause was mildly punishing, and thus, the frequency of tense stutters decreased. Uttering the word easily after the pause was rewarding, and the frequency of relaxed stutters increased.
Van Riper also developed another operant conditioning technique he called a “pullout.” He taught clients to stop in the middle of tense stutters and hold onto the sound that they were stuttering on until they could relax the muscles that were so tense they prevented saying the word. The client then finished the word in a relaxed, slow manner. By finishing the word and continuing the conversation, the client rewarded himself for an easier stutter–the behavior the client was doing just before he rewarded himself. This increased the frequency of relaxed stutters. This technique teaches a person who stutters to escape from a tense block, not by pushing or squeezing with excess tension, but by reducing the tension and finishing the word with a relaxed, nearly normal-sounding production. The reason we say that negative reinforcement is used is because a reward is given for removing the negativity of the situation. In this case, the individual himself stays in the negative situation of holding onto a moment of stuttering but then allows himself to escape only when he loosens the stutter and ends the word slowly and loosely. This act rewards the slow, relaxed production of the sound or sounds he’s stuttering on, because immediately after the slowing and relaxing, the word is released, and he can continue talking.
A third operant conditioning-based tool that Van Riper developed to modify stuttering is called a “preparatory set.” Van Riper had noticed that stutterers tended to begin their stuttered utterances by putting themselves into jammed up postures or other inappropriate speech gestures. Therefore, he taught his clients to pause momentarily before the word on which they expected to stutter and put their speech mechanisms into relaxed normal states. They then would say the word in a slow, relaxed way that would be reinforced by completion of the word and the experience of not getting stuck.
By combining these three operant-based techniques with classical conditioning to reduce clients’ negative emotions, Van Riper had a therapeutic package that effectively taught stutterers to stutter in such easy ways that they sounded like normal speakers.
Avoidance Conditioning
My stuttering in high school, described earlier, was peppered with dozens of “ums.” I usually said “um” three or four times before I even tried to say a word on which I expected trouble. Sometimes my classmates, in a curious and friendly way, would count the number of “ums” I had in a row and tell me later what my record was. Once they counted 17 “ums” as I was trying to say “Yugoslavia.” These “ums” were the result of avoidance conditioning. I first learned to use “um” as an escape behavior. When I was hopelessly stuck on a word, I could sometimes release it by quickly saying “um” and trying the word again. Then I began to say “um” even before starting a word I expected to be difficult. This sometimes prevented a stutter from occurring, and I was on my way to a serious case of avoidance conditioning. Because I rarely tried to say a difficult word without saying “um,” I never learned that I might do better without it.
There are many types of avoidance behaviors that people who stutter learn. They include avoiding speaking situations, avoiding certain words and substituting easy words for hard ones, and using extra sounds, words, or phrases, such as “um,” “well,” and “you know,” to get a running start on a difficult word. One politician I worked with cringed every time he’d see himself on television hemming and hawing and dodging difficult words, looking and sounding like a politician. He successfully learned to be open about his difficulty and stutter mildly on words he really wanted to use.
Avoidance conditioning is a big part of everyday life. The example in which you learned to escape from a long line in the bookstore by going to a cash register in the CD section can also be used to describe how avoidance conditioning may be common in your life. Perhaps after escaping from long lines at the front register several times, you might have started to avoid that register altogether and just go straight to the register in the CD section. Although many avoidance behaviors are first learned as escape behaviors, many are not. Think about why you put on your seat belt when you drive or don a helmet when you get on a motorcycle or snowboard. You may never have had an accident, but public safety spots on TV make us imagine what can happen if you don’t take these precautions to avoid brain damage or death.
Clinically, you will need to help people who stutter reduce their avoidance behaviors using a number of approaches. It is often difficult to get rid of avoidance behaviors, because the person doesn’t dare to find out what would happen if he didn’t do them. It’s like the man who was referred to a psychiatrist for habitually snapping his fingers. “Why do you do that?” the psychiatrist asked. “Because it keeps the elephants away,” the man replied. The psychiatrist then said, “But you don’t need to. There are no elephants for miles around.” “Well, then,” the man replied, “it works, doesn’t it?”
This patient might be helped if the psychiatrist could convince him to stop snapping his fingers in the office where the psychiatrist could give him protection and reassurance. This would enable the patient to see that no elephants came even though he didn’t snap his fingers for at least an hour. You can use similar techniques to help a high school student learn to tackle difficult words without saying “um.” For example, you could highly praise him for starting a feared word without saying “um” but beginning it slowly and deliberately with a “preparatory set” instead and using a “pullout” if stuttering occurs. I have always found it important to teach a client what to do when he is afraid he will stutter before teaching her what avoidance behaviors not to do.
Fortunately, some individuals who stutter will stop using avoidance behaviors without having to work directly on them. Directly improving fluency, for example, will often decrease avoidances. This is particularly true for children, who may be avoiding a few words or speaking situations, but have not developed a widespread pattern of avoidance. For them, learning how to speak more fluently will reduce their fear of stuttering enough so that they are motivated to approach all the speaking opportunities presented to them.

• During the preschool years, rapid and differential changes in a child’s body– especially in speech structures–may make it difficult for a child to coordinate the rapid movements necessary for fluency.
• A child’s cognitive development may not only compete with speech production forresources but may also provide the intellectual ability for a child to compare herself with others, which may lead to embarrassment and shame about stuttering.
• In younger children with immature brains, speech pathways may not be bufferedfrom the effects of emotional arousal, resulting in more disfluency. Some children may be especially vulnerable to the effects of emotion on speech because they can be easily emotionally aroused. Self-conscious emotions develop soon after age 2, and the emotional distress of these children in reaction to stuttering may produce increased physical tension, particularly in the larynx.
• As language develops rapidly, the increasing length and complexity of children’sutterances may sometimes exceed their speech production abilities. Selecting words, encoding phonology, planning syntax, and working out the complex prosody for an entire utterance all occur just as the child is starting to speak. It is no wonder, then, that most early stuttering occurs on the first words of sentences.
• There is some evidence that delays in language acquisition, and especially phonological development, may be associated with increased risk of persistent stuttering.
• Parents of children who stutter may, as a group, show slight tendencies to be demanding and perfectionistic. This may be the result of having a child with a speech difficulty or a manifestation of the genetic background that also results in a child having a vulnerable temperament. Research in this area is equivocal, however, with many studies finding no differences between the parents of children who stutter and those of children who don’t.
• There is no clear evidence that parents of children who stutter converse using fasterrates, more questions, more interruptions, and more long, complex utterances than do the parents of nonstuttering children. Several studies have found differences between these groups of parents, but many have not. Despite the lack of research support, many clinicians believe that it is therapeutically helpful for parents to slow their speech rates and speak in shorter sentences when talking to their children who stutter.
• There is a wealth of clinical anecdotes suggesting that difficult events in a child’slife, such as parents’ divorce or the arrival of a new baby in the family, may trigger stuttering. However, there is little empirical support for these anecdotes.
• Classical conditioning may cause stuttering to spread to many different contexts andto be consistently present rather than episodic. Operant conditioning can increase the frequency of escape behaviors. Avoidance conditioning can increase the frequency of behaviors that stutterers use to postpone or evade expected stutters. All three types of conditioning are important contributors to the establishment of secondary stuttering behaviors and are also critical to the treatment of stuttering.

1. The effect of a child’s development on fluency has been likened to the effect ofmultiple tasks for a computer. Explain this analogy.
2. It has been said that children usually do not learn to walk and talk at the sametime. What does this suggest about how motor development might affect fluency?
3. There is a high incidence of stuttering among individuals with cognitive impairment. What might this suggest about the relationship between cognition and fluency?
4. What aspects of social and emotional development might threaten fluency?
5. What evidence is there that emotional arousal might increase disfluency?
6. What is the possible connection between atypical hemispheric localization andthe effects of emotion on fluency?
7. Why would children’s speech and language development be likely to put greaterpressure on fluency than would their physical or cognitive development?
8. What aspects of parents’ behavior might put pressure on a child who is disfluent?
9. Identify several characteristics of parents’ speech that may create difficult modelsfor a disfluent child to emulate.
10. Name several life events that have been suggested to increase a child’s disfluency.
11. The communicative failure and anticipatory struggle view proposes that experiencing a communication failure may cause a child to anticipate difficulty speaking and begin to stutter as a result. What characteristic of the child may be another important factor?
12. Johnson and associates’ (1959) revised view of stuttering suggested that it resultsfrom an interaction among the following three factors: (a) the extent of the child’s disfluency, (b) the listener’s sensitivity to that disfluency, and (c) the child’s sensitivity to his own disfluency and to the listener’s reaction. Relate these factors to constitutional, developmental, and environmental factors in stuttering.
1. Record a natural speech sample from someone who stutters and analyze therelationship between the occurrences of stuttering and the linguistic level of the utterances in which they occur.
2. Develop an experimental protocol to assess the relationship between linguisticvariables and stuttering. For example, compare the variables of length of utterance, syntactic level of utterance, and phonological complexity of utterance on the likelihood of the utterance being stuttered.
3. Design a therapy activity for someone who has a classically conditioned fearof dogs. One way to do this would be to develop a hierarchy of situations that progress from Suggested Readings easy and nonthreatening to gradually more realistic encounters with dogs. At each level of the hierarchy, have the client engage in some “approach” behavior (like talking to the dog) that will counteract the old tendency to avoid dogs.
4. Study the effect of your speech rate on other people by designing and carryingout an experiment in which you vary the speed at which you talk. Record conversations in which you talk slowly for several minutes and then talk rapidly for several minutes. Measure the effect on your conversational partner’s speed of talking. You will need to practice varying your rate beforehand.
5. In the section called Speech and Language Development, research on language abilities of children who stutter is reviewed. Some studies found that children who stutter have poorer language abilities, and other studies did not. Review these studies and suggest what might be causing this disagreement in the literature.
Andrews, G. & Harris, M. (1964). The syndrome of stuttering. London: W. Heinemann Medical Books.
These authors present data from longitudinal studies of 1,000 families in Newcastle, England. The interpretation of results presents evidence that both genetic and environmental influences are at work to create stuttering. This book gives an early version of the “capacities and demands” view that stuttering is due to a lack of capacity for some aspect of speech and language processing.
Ayres, J. J. B. (1998). Fear conditioning and avoidance. In W. O’Donohue (Ed.), Learning and Behavior Therapy. Boston: Allyn and Bacon.
This chapter provides a good update of the animal learning literature on fear conditioning. Several findings appear to be relevant to stuttering. First, some responses to fear may be part of an animal’s hardwiring (perhaps laryngeal tension is a natural response to fear of speaking in some individuals). Second, some individual animals freeze in response to fear rather than learn an effective coping response (are some children who begin to stutter more predisposed to tense blockages than others?). Third, for a fear-conditioned response, reducing fear without teaching a new response to fear leaves the animal vulnerable to relapse (it may be important to reduce stutterers’ fear of stuttering as well as teaching them a coping skill they can use when fear is present).
Bernstein Ratner, N. (1997). Stuttering: A psycholinguistic perspective. In R. Curlee & G. Siegel (Eds.), Nature and Treatment of Stuttering: New Directions (2nd ed.). Boston: Allyn & Bacon.
This is an insightful review of the many connections between language and stuttering. The author’s background allows her to use linguistic theories and evidence from child language studies to discuss how language influences the loci of stuttering in speech, how parent-child interactions may affect stuttering, how language development may be important in stuttering onset, and the role of feedback on speech, language, and stuttering development.
Bloodstein, O., & Ratner, N. (2008). Inferences and conclusions. In
Bloodstein, O., A Handbook on Stuttering. Clifton Park, NY: Thompson-

Delmar Learning.
This chapter presents the communicative failure and anticipatory struggle view of stuttering onset. Bloodstein musters the evidence he has summarized in earlier chapters of this handbook to argue convincingly that stuttering develops from an interaction between the child and his environment.
Crystal, D. (1987). Towards a “bucket” theory of language disability: Taking account of interaction between linguistic levels. Clinical Linguistics and Phonetics, 1, 7–22.
A theoretical discussion of interaction among levels of speech and language, with an illustrative case of a child whose stuttering increases when language demands are greater. The article makes a clear argument for the influence of speech and language development on stuttering.
Johnson, W., et al. (1959). The onset of stuttering. Minneapolis, MN:
University of Minnesota Press.
This book presents extensive data on parents’ perceptions of the onset of their child’s stuttering, compared with other parents’ perceptions of their child’s normal disfluency. Johnson eloquently lays out his view of stuttering as the product of an interaction between the child’s disfluency, his sensitivity, and the listener’s reactions.
Kagan, J., Reznick, J. S., & Snidman, N. (1987). The physiology and psychology of behavioral inhibition in children. Child Development, 58, 1459– 1473.
This article discusses the findings that behaviorally inhibited children show high levels of laryngeal tension. Neurophysiological mechanisms are also discussed as well as possible genetic and environmental contributions. Recommended for those interested in the hypothesis that behavioral inhibition may be a component in some stuttering.
Paden, E. P. (2005). Development of phonological ability. In Yairi, E., & Ambrose, N., Early Childhood Stuttering. Austin, TX: Pro-Ed.
This chapter focuses on the phonological development of children who stutter with particular emphasis on comparisons between children who recover without intervention and those who persist in stuttering. The author brings to light several aspects of her research that are intriguing puzzles for future researchers to solve.
Watkins, R. V. (2005). Language abilities of young children who Stutter. In Yairi, E. & Ambrose, N. Early Childhood Stuttering. Austin, TX: Pro-Ed.
Although current evidence reviewed in this chapter suggests that language abilities of children who stutter and those who don’t are similar, language factors appear to play an important role in stuttering. The author discusses several interesting relationships between language and stuttering, including the role of language factors in the occurrence of stuttering in an utterance and the finding that early onset of stuttering is often associated with advanced language skills.
Research Findings on Developmental,
Environmental, and Learning Factors in Stuttering

Physical and Motor Skill Development
Speech and Language Development
Delayed and Deviant Speech and Language Development
Cognitive Development
Environmental Factors
Speech and Language Environment
After studying this chapter, readers should be able to discuss research findings on developmental, environmental, and learning factors in stuttering in greater depth
Like Chapter 3, this chapter describes in detail the research studies that provide evidence about factors important in understanding stuttering that were summarized in the previous chapter–in this case, Chapter 4. Note that not every topic discussed in the previous chapter is covered here, because in some cases, the research is minimal.
In the main section of this chapter, I suggested that developmental factors may influence stuttering in young children. Here, I will make the case that speech motor skill (motor commands) development in all children puts great demands on neural resources needed for fluent speech, and it may present a special challenge for children who stutter because of differences in brain structure and function.
One problem posed by physical development in children is rapid change in the vocal tract between ages 2 and 5 years. During this time, structures in a child’s head, neck, and torso undergo their most accelerated growth; moreover, different structures grow at different rates (Kent & Vorperian, 1995, 2007). As children’s speaking mechanisms change day by day–the shape, size, and biomechanical properties of muscles and bones are different today from the way they were yesterday–children somehow manage to continue to produce intelligible speech. Callan, Kent, Guenther, and Vorperian (2000) propose that children maintain a stable speech output in the face of almost daily changes in their speech structures by using feedback to continuously update the motor commands they send to their muscles to produce specific sounds. What their brains told their muscles to do yesterday must be adapted to the new size, shape, and biomechanical properties of the vocal tract today. Auditory feedback, integrated with proprioceptive and other muscle feedback, helps children discover errors in their motor commands and adjust the commands to the new dimensions.
The theory behind this hypothesis may be important for understanding theories of stuttering discussed in Chapter 6 (e.g., Neilson & Neilson, 1987, 1988), so I will spend a little extra time explaining it. A key assumption about learning to speak is that from the day he is born, a child is learning the relationship between what his brain tells his muscles to do (motor commands) and what sounds come out of his mouth (perceptual target). He begins this learning with his earliest cries and babbles. As he begins to attend to the speech around him, he develops a sensory memory of those sounds and then practices the motor commands needed to produce them. In doing so, he develops “sensory-motor neural maps” that he uses to generate these motor commands in a reliable way (Kent & Vorperian, 1995, 2007). These maps are sometimes called “inverse internal models” of the relationship between the sensory-motor target that the child wants to hit and the motor plan needed to hit it. They are called “inverse” because the child’s brain selects the target to hit (the sound he wants to make) and instantly calculates how to hit it with movements of articulators, larynx, and breathing apparatus, given their current positions. As the child’s vocal tract changes and matures, he maintains and updates the relationships between motor commands and acoustic output from the information provided by the auditory feedback of his own speech, as well as feedback from the touches and movements of his articulators involved in producing that speech.
Evidence from brain imaging studies presented in Chapter 2 (Cykowski, Fox, Ingham, Ingham, & Robin, 2010; Sommer, Koch, Paulus, Weiller, & Buchel, 2002) suggests that bidirectional pathways connecting sensory integration areas with speech motor planning areas are less dense in individuals who stutter. Thus, the pathways that were designed to build and update the sensory-motor models of the sounds and words they want to say may be inefficient in children who stutter. These pathways not only help the child as he learns to speak accurately during development, but they are used for every utterance. When these children plan to speak and need quick access to the stored inverse internal models, this access may be slow in coming, and stuttering may result.
This section contains research and commentary about the phenomenon that the first appearance of stuttering frequently happens when a child is rapidly developing speech and language.
Many writers have eloquently commented on the connection between the onset of stuttering and language acquisition. Peggy Dalton and W.J. Hardcastle (1977), for example, commented that “it is tempting to see the ever-increasing demands on linguistic competence and articulatory proficiency as major factors in the onset of some disfluency.” Joseph Sheehan (1975) said, “The age of onset of stuttering is consistently related to certain stages in the developmental sequence. Most notably, the ‘period of resonance,’ or high readiness in language learning…is also the period during which stuttering develops and flourishes” (p. 142). Andrews and colleagues (1983) pointed to the demands placed on speech by rapidly developing language, noting that “stuttering [has] a maximal frequency of onset at a time when an explosive growth in language ability outstrips a still-immature speech-motor apparatus” (p. 239).
Many studies have found that greater length and complexity of language is associated with more stuttering. This is a developmental factor because stuttering may first appear when–in those children predisposed to stutter–their language matures and they use longer and more complex utterances. For example, research on natural conversational speech of children who stutter has shown that more complex utterances contain more stuttering (Brundage & Bernstein Ratner, 1989; Gaines, Runyan, & Meyers, 1991; Logan & Conture, 1995; Yaruss, 1999). Some research suggests that utterance length may have a greater effect on stuttering than does complexity (Logan & Conture, 1995; Wilkenfeld & Curlee, 1997; Yaruss, 1999). Experimental studies, in which children were asked to produce both more and less complex utterances, show that both stutterers (Bernstein Ratner & Sih, 1987; Stocker & Usprich, 1976) and nonstutterers (Gordon, Luper, & Peterson, 1986; Haynes & Hood, 1978; Pearl & Bernthal, 1980; Yaruss, Newman, & Flora, 1999) increase their disfluencies as language complexity is increased. Unfortunately, there is little longitudinal research that directly bears on the question of how and when emerging language is associated with normal disfluency or stuttering.
One of the few descriptive studies of this phenomenon was Norma Colburn’s analysis of the disfluencies of four nonstuttering children using data originally gathered by Lois Bloom for her work on normal language development. Published reports of her analysis (Colburn & Mysak, 1982a, 1982b) suggested that these children’s normal disfluencies did not emerge when they first learned a new language construction but as they began to master it and started using it regularly. Explanations suggested for this result include the possibility that a child who has not completely automatized the use of a new construction allocates fewer resources than are necessary for its production (Kent & Perkins, 1984) or that when a child masters the new construction he produces it at an increased rate, thereby straining capacity (Starkweather, 1987).
A single-case study by Frank Wijnen (1990) was used to explore the relationship between syntax acquisition and normal disfluencies. Weekly speech samples were obtained from a boy from age 2 years, 4 months to 2 years, 11 months. The number of repetitions, revisions, and incomplete phrases was assessed in relation to the length and complexity of utterances. It was reported that disfluencies were randomly distributed initially, but eventually clustered on function words and sentence-initial words and then declined. Although speech rate was not measured (increased rate might have accounted for some of the increase in the child’s disfluencies), the number of disfluencies was not highly correlated with length of utterance. Instead, Wijnen concluded that the eventual decline in the child’s disfluencies was associated with his development of a routine type of sentence (pronoun + verb + some other word) and that learning this routine involved so much of his processing capacity that speech production was shortchanged and initially caused the disfluencies. This preliminary study needs to be followed up with many more cases to test the hypothesis that the process of first learning to make sentence productions more automatic through routinization of several sentence types is related to increased normal disfluency. With a larger sample size, multiple regression analyses could be used to determine which of many possible factors best predict instances of or increases in disfluency.
Having looked at the evidence of an association between increasing sentence complexity and/or sentence length and the frequency of stuttering in children, let’s consider what may be specifically interfering with fluency as the child acquires language. I’ll begin by recounting what the child acquires during this period of time. Between ages 2 and 3, a child’s vocabulary jumps from 50 to well over 500 words; in fact, toward the end of this year, five to seven new words may be learned each day (Studdert-Kennedy, 1987). At the same time, the child’s single-word utterances develop into successive single-word pairs with sentence-like intonations and durations, then to multi-word sentences (Branigan, 1979). Thus, the child’s speech graduates from a simple “syllable-timed” prosody for single words to complex prosodic rhythms that span multiple words (Allen & Hawkins, 1980). As the child expands sentences, he also overhauls his language storage system. At first, his shelves are stocked with whole words in the form of articulatory routines or gestural patterns; then, he must change strategies and begin to stock, not whole words, but segments that can be combined in various ways to form a multitude of words (Kent, 1985; Nittrouer, Studdert-Kennedy, & McGowan, 1989; Sternberger, 1982). During these same early preschool years, the child also progressively learns active, negative, and passive constructions as well as present, future, and past tenses. At the same time, he increases the length and linguistic complexity of his sentences together with the rate of his utterances as he increasingly tries to synchronize the rates and rhythms of his speech with those of his family, with whom he has a growing urge to communicate.
This huge array of language and speech production tasks is a challenge even to the fluency of nonstuttering children. Normal disfluencies of children also increase from ages 2 to 4, peaking when they tackle the task of producing long, complex sentences (Ito, 1986). It is not surprising, then, that children who are predisposed to stutter because of genetic or congenital factors tend to begin stuttering during this same period.
I would like to move from what may interfere with a child’s fluency to consider how it may occur. Imagine that a child’s brain has a limited amount of space that is chock full of highly interconnected networks of neurons (Kinsbourne & Hicks, 1978). Because the child’s brain is immature but tightly packed, neural networks may not be well insulated from one another. The insulation of axons is provided by myelin sheaths, fatty protective coatings that develop slowly throughout childhood. Thus, in a younger child, myelinization is incomplete, allowing “cross talk” among axons of different neural networks, creating interference in the transmission of information. In Chapter 3, I described early speculations by Karlin (1947) about delayed myelinization as a cause of stuttering and more recent findings of Cykowski and colleagues (2010) that delayed myelinization in key neural pathways may result in disruption of appropriate sensory-motor integration for fluent speech.
Another challenge to the child developing language may be introduced by different neuronal groups maturing at different rates. Neuronal groups working on vocabulary may be more efficient than those working on syntax, or vice versa. In addition, structures–such as the corpus callosum–that link prosodic functions of the right hemisphere with syntactic functions in the left may be slower to develop than other structures involved in speech-language production. All these differences may account for why a young child may have trouble synchronizing the tasks required for speech, some of which are simultaneous while others are sequential. As a child grows, however, his brain gains more functional space with expanded neural networks that can function more independently from one another. As different groups of neurons mature and become more efficient, they can synchronize their actions more easily. Thus, when the child becomes a teenager, he can rub his tummy and pat his head with less disruption of either task.
Returning now to the realm of speech and language production, when a young child is at the one-word stage of language development, the amount of simultaneous activity in different neural networks may be relatively small. As I indicated earlier, the child’s task is to select a word from a small shelf of whole words and then produce it with simple prosody. But when the child moves to the two-word stage and beyond, he must select several words to make each sentence, and his chosen words must be assembled from a large storehouse of smaller segments rather than a storehouse of whole words. He must also work out a grammatical plan for the sentence and align it with a complex rhythm that spans the entire sentence. Some of the planning for the later parts of the sentence is going on at the same time he is beginning to produce the first words of the sentence, putting greater stress on developing cognitive factors such as memory. Clearly, these simultaneous but different tasks, involving interconnecting neural networks that are not fully mature, will sometimes interfere with each other. Think of the errors we all make when we try to do too many different things at the same time. The hesitations, pauses, and repetitions in the speech of normal children who are learning to talk in sentences may be the equivalent of the stops and starts and confusion that children show when simultaneously performing tasks that interfere with each other.
Delayed and Deviant Speech and Language Development
When assessed on such measures as the ages when they said their first words or uttered their first sentences, the size of receptive vocabulary, mean length of utterance, and expressive and receptive syntax, children who stutter often score lower than their nonstuttering peers (Andrews & Harris, 1964; Arndt & Healey, 2001; Bernstein Ratner & Silverman, 2000; Berry, 1938; Darley, 1955; Kline & Starkweather, 1979; Murray & Reed, 1977; Ntourou, Conture, & Lipsey, 2010; Okasha, Bishry, Kamel, & Hassan, 1974; Wall, 1980; Westby, 1979; Williams, Melrose, & Woods, 1969). However, other studies have not found language differences (e.g., Johnson, 1955; Miles & Ratner, 2001; Peters, 1968; Seider, Gladstien, & Kidd, 1982; Watkins, Yairi, Ambrose, & Grinager, 1999).
Young children who stutter have also been shown to have difficulty achieving ageappropriate speech. In the clinic, we often observe children in the early stages of stuttering who have multiple articulation or phonological errors and speech that can be difficult to understand. Research has repeatedly confirmed the finding that stutterers have roughly two and a half times the incidence of articulation disorders as that found in same-age nonstutterers (Andrews & Harris, 1964; Berry, 1938; Bloodstein, 1958; Bloodstein & Ratner, 2008; Kent & Williams, 1963; Williams, Silverman, & Kools, 1968). Nevertheless, several studies have found no differences in the articulation abilities of children who do and do not stutter (Ryan, 1992; Seider, Gladstien, & Kidd, 1982).
At least two excellent, critical reviews of the research on language, phonology, and stuttering have been published. Nippold (1990) suggests, for example, that research does not clearly support the hypothesis that children who stutter are also likely to have language or articulation difficulties. Rather, she proposes that there may be subgroups of children who stutter who have language or articulation problems related to their stuttering. Bernstein Ratner (1997) and Bloodstein and Ratner (2008) concurred and noted that the differences found between groups of children who stutter and children who don’t stutter have been very subtle. They suggested that future research should use more sophisticated tests of language and phonology and look for subgroupings, not only in children but also in adults who stutter. Although not strictly a review of the literature, Watkins (2005) summarizes this research and gives an excellent overview of the many findings related to language and stuttering in the Illinois Stuttering Research Program. Her chapter suggests that there is no strong evidence that children who stutter have language delays or language difficulty; on the contrary, many children in their studies showed advanced language abilities near the onset of stuttering.
Findings of deficits in articulation and language performance, at least in some children who stutter, can be interpreted in several ways. Some authors have suggested that children who have articulation or language difficulties will start to believe that speaking is difficult. Their anticipation of such difficulty is hypothesized to lead to hesitation and struggle and then to stuttering (Bloodstein, 1995, 1997; Bloodstein & Ratner, 2008). An alternative view is that stuttering, language disorders, and articulation errors all result from a common deficit, which might be passed on genetically. Because specific regions and pathways of the brain are responsible for speech and language-related functions, delayed development of (or damage to) these areas may result in language, articulation, or fluency problems in any combination. Small differences in how the brain processes such functions could tip the balance toward any of these disorders.
The literature on disfluencies of language-impaired children who are not considered stutterers provides support for a slightly different hypothesis about the relationship between language deficits or delays and stuttering. Several studies of languageimpaired children have found that they–or a subgroup of them–evidence high frequencies of the types of disfluencies that are seen more often in children who stutter than in normal children, even though they wouldn’t be considered stutterers (Boscolo,
Ratner, & Rescorla, 2002; Hall, Yamashita, & Aram, 1993; Hodge, Rescorla, & Ratner, 1999). These authors speculate that the excess disfluencies result from difficulties in formulating and executing utterances just beyond the limits of their language abilities. Thus, when language resources are strained, fluency is sacrificed in order to meet the demands of language production. Perhaps these children need extra time to meet these demands, and part- and whole-word repetitions provide the needed time, or this may be what results when a speech production system is spinning its wheels, waiting for elements of the language plan to be ready for execution. These findings suggest that the association between stuttering and language delay or deficit may emerge from the demands of normal development on weak language formulation/production systems and result in high frequencies of stuttering-like disfluencies. Other factors, such as a vulnerable temperament, environmental pressures, or a traumatic life event, may turn these disfluencies into real stuttering.
In addition to the evidence that children who stutter have a higher incidence of language or articulation problems, researchers are finding evidence that articulation or language delays and disorders may be related to whether or not a child recovers from stuttering. This is reflected in the research of Conture, Louko, and Edwards (1993), Paden (2005), St. Louis (1991), and Yairi, Ambrose, Paden, and Throneburg (1996), whose findings suggest that children who stutter and also have phonological or language differences are more likely to persist in stuttering or take longer in treatment. As part of a large ongoing study of children identified and assessed close to the onset of stuttering, Yairi and colleagues (1996) found that children who later recovered from stuttering without treatment had higher scores on the Preschool Language Scale (Zimmerman, Steiner, & Pond, 1979) than the children who did not recover. It is noteworthy, however, that the children who later recovered as well as those whose stuttering persisted scored above the norms on this test. A more recent study of an extended cohort of these same children (Watkins, Yairi, & Ambrose, 1999) did not replicate the study’s initial finding. Specifically, lexical, morphological, and syntactic analyses of the cohort’s spontaneous language close to the time they began to stutter showed no differences between recovered and persistent groups of children. However, in looking at the phonological development of both groups near onset of their stuttering, Paden, Yairi, Ambrose, and Grinager (1999) found that the children who recovered from stuttering scored significantly higher on the Assessment of Phonological Processes-Revised (Hodson, 1986) than did those who did not recover. Thus, it appears that among factors measured near the onset of stuttering, a child’s phonological status, but not expressive language status, may predict recovery. Children with delayed phonology are at risk for stuttering to persist.
The studies reviewed here are not directly concerned with cognitive development but do show the link between cognitive abilities and stuttering. First, there is strong evidence that people with cognitive deficits, especially when deficits are relatively severe, have a high incidence of stuttering (Van Riper, 1982). An explanation has been suggested by Starkweather (1987), who noted that developmentally delayed individuals are slower in their overall acquisition of speech and language. Their extended period of acquisition may make them more vulnerable to speech breakdown because competition between language acquisition and motor speech production for limited neurological resources occurs over a relatively long period of time.
Individuals who have had traumatic brain injury, which usually affects cognitive functions such as memory and attention, also have an increased risk for fluency disorders (Jokel, De Nil, & Sharpe, 2007; Theys, van Wieringen, & De Nil, 2008). This may occur for more than one reason. In the first place, typically rapid and complex speech and language production depend on fully functioning perception, attention, working memory, and executive functions. When these processes are compromised, breakdowns in spoken language are likely to result. As an example, consider the effect of a faulty working memory on rapid retrieval of vocabulary or syntax. If some components of language are mistimed in relation to others, repetitions of words or syllables may result, just as an engine with an unsteady fuel supply will stop and start, stutteringly.
Yet another link between cognition and stuttering is found in a study by Yairi and colleagues (1996) indicating that poorer cognitive skills are associated with lack of ability to recover from stuttering. Of the study’s 32 children who began to stutter, 12 continued to stutter for 36 months or more. The two groups of stutterers, those who recovered and those who did not, were compared with a control group of nonstuttering children on an intelligence test–the Arthur Adaptation of the Leiter International Performance Test (Arthur, 1952). The group of children who continued to stutter scored significantly lower than the nonstuttering control group, although their mean score was not below the norm for the test. However, the children in the recovered group were statistically identical to the control group. Thus, some abilities associated with cognition may be related to a neural resilience allowing recovery from stuttering. In other words, children with slightly higher cognitive functioning may have the extra resources needed to reorganize their speech and language processing, allowing them to develop a workaround for the problem causing them to stutter.
In the 1930s and 1940s at the University of Iowa, Wendell Johnson developed the “diagnosogenic” theory of stuttering (1942). This theory, which is described more fully in the next chapter, proposes that a child’s parents misdiagnose normal disfluencies as stuttering. Their reaction to the “stuttering” then causes the child to try to avoid these normal interruptions of speech and struggle in a way that becomes real stuttering. Johnson’s diagnosogenic theory generated a great deal of research on parents of stutterers. Were they different from the parents of nonstutterers? Were they unusually critical? Did they have unreasonably high standards of speech?
One of the first studies of parents of stuttering children was conducted by John Moncur (1952). He interviewed the mothers of both stuttering and nonstuttering children about their parenting practices and concluded that mothers of stuttering children tended to be more critical, more protective, and more domineering toward their children than the mothers of nonstuttering children. Not long afterward, Frederick Darley, a student of Wendell Johnson, investigated the attitudes of stutterers’ parents in more detail. Using interview techniques based on Alfred Kinsey’s studies of sexual behavior, Darley (1955) questioned the parents of 50 stutterers and 50 nonstutterers. Although there was a great deal of overlap in the attitudes of both groups, parents of children who stuttered had significantly higher standards and expectations, particularly with regard to speech. They had greater sensitivity to speech deviations and believed in early intervention for disfluencies, and their overall drive and need for domination was greater as well.
Johnson and his research associates (1959) expanded Darley’s study to the parents of 150 stuttering and 150 nonstuttering children. Again, there was much overlap between both groups, but parents of children who stutter were reported to be more perfectionistic and to have higher standards of behavior than the parents of children who did not stutter. These studies by Moncur, Darley, Johnson, and others were largely responsible for the widespread belief that parents are a key factor in precipitating the onset of stuttering. Parents could transmit a culture’s “competitive pressure for achievement or conformity,” which may be the environmental factor most likely to be linked with stuttering (Bloodstein, 1987).
Let us digress for a moment from the Iowa studies that found competitive pressures to be common in the homes of stutterers. Different results were found in England. Gavin Andrews and Mary Ann Harris (1964), whose work was described in Chapter 2, collected and analyzed data from families in Newcastle, England. Andrews and Harris compared the medical and home visit records of parents of stutterers with those of parents of nonstuttering children and concluded that both groups of parents, most of whom were mothers, were generally similar in personality but differed in some key traits. The parents of stutterers were lower in intelligence, had poorer school records when they were younger, had poorer work histories, and provided poorer housing for their children. There was no evidence, however, that they criticized or pressured their children. This finding is a far cry from the reports of excessively high standards in the homes of children who stutter in Iowa.
Why are these results so different from those of the Iowa studies? There may be many reasons, but two come readily to mind. First, stuttering may emerge in children under many types of stress. In industrial England, the greatest stress may have come from social and economic disadvantages, but the greatest stress in Iowa may have been the high standards of upwardly mobile parents. Second, these different results may reflect differences in these researchers’ expectations and theoretical biases of several decades ago. Americans tend to believe that everyone is created equal, and American researchers, especially those conducting studies in the heartland of the United States during the 1950s, would be predisposed to look for causes of stuttering in the environment rather than the child’s heredity. In contrast, many British people at that time may have believed that inheritance plays a major role in determining one’s life outcomes, and researchers in the United Kingdom would have been inclined to consider parents’ intelligence and social class as likely causes of stuttering.
The hypothesis that lower class homes provide stress that may precipitate stuttering did not originate in England. John Morgenstern (1956), who investigated stuttering in Scottish children, found that stuttering was more prevalent in the lower class homes of skilled manual, weekly wage earners compared to homes in the Iowa studies. This, however, was a social stratum in Scotland that was upwardly mobile and may have expressed families’ ambitions through high speech standards for their children. Here, we have a combination of forces if Morgenstern’s hypothesis was correct. The stress of lower class homes lies not in their deprivation but in the cultural pressure to perform well and rise above humble beginnings.
Studies of parents of children who stutter compared to parents of children who don’t present mixed results. Some have reported that parents of children who stutter are more rejecting or anxious than are parents of children who do not stutter (Flugel, 1979; Zenner, Ritterman, Bowen, & Gronhovd, 1978), but others have found either small or no differences between the two groups of parents (Goodstein, 1956; Goodstein & Dahlstrom, 1956). In his thorough review of the home environments of children who stutter, Yairi (1997b) concluded that the mix of diverse findings boils down to the likelihood that children who stutter grow up in unfavorable environments –homes that may stress children. But he also noted that even though many studies suggest that parents of children who stutter may be somewhat anxious, overprotective, socially withdrawn, and prone to negatively evaluate their children, there is no evidence that these parental tendencies cause stuttering. Yairi went on to point out that since a child’s risk for stuttering is often inherited, the parents of these children may themselves stutter, have stuttered in the past, or had contact with other family members who stuttered. Thus, their negative traits may be a result of their own experiences with this disability.
Speech and Language Environment
Several clinical researchers have examined parent-child conversational interactions, seeking to determine if parents of children who stutter talk to their children differently than parents of nonstuttering children. A review of many such studies was published by Nippold and Rudzinski (1995). In the following paragraphs, we highlight some of the important studies in this area. Susan Meyers and Frances Freeman (1985a, 1985b) compared the speech of mothers of stuttering children with that of mothers of nonstutterers. They found that mothers of stutterers spoke more rapidly than did the mothers of nonstutterers. This may be critical, since a mother’s high speech rate may encourage a child to try to speak faster than his optimal speed (e.g., Jaffe & Anderson, 1979). The possibility that rapid speech rates may lead to stuttering is consistent with Johnson and Rosen’s (1937) finding that adult stutterers were more likely to stutter when they spoke more rapidly than their habitual rates. Children who stutter may be even more vulnerable to fluency breakdowns during rapid speech than adults who stutter by virtue of the fact that children’s natural rates of speech are slower and their temporal coordination less than those of adults (e.g., Kent, 1981). Also, remember from Chapter 2 that Kloth, Janssen, Kraaimaat, and Brutten (1995), Kloth, Kraaimaat, Janssen, and Brutten (1999) found that children at risk for stuttering who developed stuttering spoke faster than those at risk who didn’t develop stuttering.
However, several studies subsequent to those of Meyers and Freeman (1985a, 1985b) failed to find speech rate differences between parents of children who stutter and parents of children who don’t stutter. For example, Kelly and Conture (1992) found no differences in the speaking rates of mothers of these two groups of children, Kelly (1994) found no differences in the rates of fathers of the two groups, and Yaruss and Conture (1995) found no differences in the articulatory rates (the rate at which each individual phrase is spoken, in contrast to “speaking rate,” which includes pauses between phrases) between mothers of stuttering children and mothers of nonstuttering children. However, the latter researchers did find a significant correlation between children’s severity of stuttering and parent-child differences in speaking rate; greater differences in parent-child speech rates were associated with more severe stuttering in the children. These results could have been obtained if more severely stuttering children talk more slowly than other children and their parents have speech rates similar to other parents in the study. It is also possible that greater parent-child differences in rate make a child who already stutters more severe.
In general, earlier studies found differences between parents’ speaking rate but later ones didn’t. Why might this be so? Zebrowski (1995) discussed this disparity in her review of the conversational patterns of families of children who stutter. She suggested that differences in measurement techniques and the fact that Meyers and Freeman (1985a, 1985b) had a larger number of severe stutterers in their sample might account for the differences in the studies. We think it is possible, also, that parents of children just beginning to stutter have become increasingly aware of the importance of speaking slowly because of publicity aimed at stuttering prevention. Such parents may try to speak more slowly than is typical for them while they are under the scrutiny of clinical researchers, thus adding to the likelihood that more recent studies may not find differences in the speaking rates of parents of stutterers and nonstutterers.
Another suspected parental stress, in addition to rapid speech rates, is the frequency with which parents interrupt their children. One of Meyers and Freeman’s reports (1985a) presented some unexpected evidence about interruptions. The mothers of both stuttering and nonstuttering children interrupted most frequently when a child was disfluent. It seems possible that such parental interruptions, some of which may have been elicited by the child’s disfluencies, may in turn elicit changes in the child’s speech. Some children might increase tension and rate, thereby developing the struggled behaviors of stuttering. Others might suppress disfluencies to avoid interruptions and eventually be “taught” by parents not to be disfluent.
In a later study, Kelly and Conture (1992) found no significant differences in the interruptions of mothers of stuttering children and those of mothers of nonstuttering children. However, a closer inspection of their data revealed a correlation between the duration of “simultalk” (one person talking at the same time another is talking) of the mothers of children who stutter and the severity of their stuttering. Thus, mothers of more severe stutterers did more simultalk when their children were talking than did mothers whose children stuttered less severely. In a later study of fathers, Kelly (1994) found no differences in the interruptions of fathers of children who stutter and those of fathers whose children don’t stutter. Moreover, the correlation between these fathers’ simultalk and severity of stuttering was not significant.
Another variable of children’s speech and language environments that has been studied is the extent to which parents ask questions. Meyers and Freeman (1985a) found no significant difference in the number of questions asked by mothers of children who stutter compared to that of mothers of children who do not. Langlois, Hanrahan, and Inouye (1986), however, did find significant differences when making a similar comparison. Langlois and Long (1988) then conducted an experimental treatment of a 4-year-old who stuttered, in which the mother was taught to reduce the number of questions she asked, among other changes. After 16 treatment sessions in which the mother had markedly reduced her number of questions and given her child more speaking turns, the child no longer stuttered. This finding is especially interesting because it is tempting to assume that asking questions results in more stuttering. However, subsequent studies tested this assumption and failed to support it. In a study of eight stuttering children in conversations with their parents, Weiss and Zebrowski (1992) found that the children stuttered less when they answered questions than when they made assertions. This appeared to be related to the fact that questions were often answered with brief responses, but assertions were often longer responses. A more direct test of the effect of parents asking questions was carried out by Wilkenfeld and Curlee (1997), who used a single-subject ABAB design to vary an adult’s verbal behavior (questions versus comments) in conversations with a child who stuttered. Their results with three children who stuttered demonstrated that stuttering did not appear to be related to whether the adult asked questions or commented but was more likely to occur in either condition when the child’s utterances were longer.
It is becoming evident that findings about speech rates, interruptions, and questions in the conversations of parents with their children who stutter are equivocal. But what about the linguistic complexity of their speech? Miles and Ratner (2001) conducted a study of the overall complexity of the language of parents of children who stutter, gathering samples of conversations from 12 mother-child pairs involving stuttering children and 12 involving fluent children. All children were between 27 and 48 months of age, and the stuttering children were within three months of the onset of their stuttering. Mothers’ utterances were assessed for syntactic complexity, lexical diversity and rarity, and mean number of utterances per turn. No significant differences between the mothers of the stuttering children and the mothers of the fluent children were found.
Two recent studies have taken a different approach to examining the effects of parents’ verbal behavior on children’s stuttering. The complexity of mothers’ language was examined in relationship to whether or not their children recovered from stuttering. Kloth and colleagues (1999) studied 23 children who stuttered, 16 of whom had recovered after four years and seven of whom had not. The complexity of the mothers’ language was measured in terms of mean length of utterances in words in conversations with their child, both before and immediately after the onset of stuttering. They found that the language of the mothers of children who persisted in stuttering was significantly more complex than that of the mothers of children who recovered both before their children began to stutter and again after stuttering began. In another study of persistent and recovered stutterers, Rommel, Hage, Kalehne, and Johannsen (2000) assessed the complexity of the language of 71 mothers soon after their children had begun to stutter, rather than before stuttering began. They followed these children for three years and found that one of the more powerful predictors of whether or not a child would recover were “the linguistic demands to which the child is exposed” (p. 181). More specifically, these researchers found that the more complex the mother’s syntax (mean length of utterance or MLU) and the greater number of different words she used in talking to her child, the more likely that her child would not recover over the following three years. The language abilities of the children had no predictive value (Rommel et al., 2000). The findings of these two studies support the hypothesis that the language environment of a child who stutters may influence recovery.
Nippold and Rudzinski’s (1995) review of research on parents’ speech and children’s stuttering, covering studies published through the mid-1990s, makes it clear that there was only weak support for the hypothesis that parents of children who stutter have more demanding verbal interaction styles than do other parents. However, the findings by Kloth and colleagues (1999) and Rommel and colleagues (2000) about recovered and persistent stuttering in children suggest that earlier researchers were asking the wrong questions. Asking if the parents of children who stutter talk faster, interrupt more, ask more questions, or use more complex language than do other parents suggests that earlier studies had assumed that such parent behaviors causes stuttering in otherwise normal children. However, given the strong evidence cited in Chapters 2 and 3 that stuttering has a neurophysiological basis, it is unnecessary to look for a cause of stuttering in demanding parental speech patterns. It may be that parents of children who start to stutter have relatively typical speech and language patterns when talking to their children. The problem may be that these typical patterns don’t particularly nurture the growth of fluency in children just starting to stutter, and thus, these children persist in stuttering. On the other hand, parents of children starting to stutter who simplify their language and slow their speech rates may facilitate their child’s recovery from stuttering.
This ends the summary of the research findings on developmental, environmental, and learning factors associated with stuttering. I would like to remind the reader that the most recent article or book cited in this chapter is 2010. More up-to-date findings will be uncovered by using databases such as PsycINFO, PubMed, and Google Scholar.
Theories about Stuttering

Theoretical Perspectives about Constitutional Factors in Stuttering
Stuttering as a Disorder of Brain Organization
Stuttering as a Disorder of Timing
Stuttering as Reduced Capacity for Internal Modeling
Stuttering as a Language Production Deficit
Stuttering as a Multifactorial, Dynamic Disorder
Theoretical Perspectives on Developmental and Environmental Factors
Diagnosogenic Theory
Communicative Failure and Anticipatory Struggle
Capacities and Demands
An Integration of Perspectives on Stuttering
A Two-Stage Model of Stuttering
A Perspective on Primary Stuttering
A Perspective on Secondary Stuttering
Two Predispositions for Stuttering
Interactions with Developmental Factors
Interactions with Environmental Factors
Interactions of Anomalous Neural Networks with Environmental Factors
Interactions of Temperament with Environmental Factors
Implications for Treatment
Accounting for the Evidence
Stuttering Occurs in All Cultures
Stuttering is a Low-Incidence Disorder
Stuttering Does Not Begin with the Onset of Speech
Stuttering Sometimes Begins with Tense Blocks, but Often with Repetitions Not All Stutterers Have Relatives Who Stuttered
Stuttering Appears as Repetitions, Prolongations, and Blocks
Stuttering Is More Common in Boys Than in Girls
In Many Children, Stuttering Starts as Mild and Develops into a More Severe Form
Many Conditions Reduce or Eliminate Stuttering
Stutterers Often Have Poorer Performance on Sensory and Motor Tasks
After studying this chapter, readers should be able to:
• Explain what a theory is and what hypotheses are
• Identify five perspectives about constitutional factors in stuttering
• Identify three perspectives about developmental and environmental factors instuttering
• Describe the author’s integrated view of stuttering
Theory: An explanation of some phenomenon; regarding stuttering, a theory might explain why some people stutter and others don’t
Hypothesis: A specific and testable proposition derived from a theory
Hemispheric dominance: In general, the phenomenon that one hemisphere of the brain (left or right) takes the lead or is stronger for a particular function. In the context of this chapter, hemispheric dominance refers to the fact that the left side of the brain is usually more specialized for speech and language than the right side
Inverse internal models of the speech production system: A concept about how your brain functions when you learn to talk as a child and as you are talking as an adult. The basic idea is that as a child hears speech in his environment, he stores auditory images of the sounds and words. During babbling, he learns how to send motor commands to his muscles to make those sounds and words These connections create the internal model for speech production. They are called inverse because they start out as the auditory images or targets but get “inverted” to become the motor commands needed to hit those auditory targets
Sensory targets: These are essentially the auditory targets mentioned in the last definition, but actually the targets contain not only auditory information but also information about the movements of the articulators (kinesthetic and proprioceptive information)
“Covert repair” hypothesis: An explanation of stuttering as the result of the brain’s stopping production of speech when it detects an error in the plan that

the brain has made to produce a word
Multifactorial, dynamic disorder: Stuttering may be seen as multifactorial because many factors (e.g., genetic, emotional, cognitive, social, environmental) interact to create it. It is also dynamic because the overt signs of stuttering are seen as surface manifestations of an ever-changing neurophysiological process underlying the disorder
Diagnosogenic theory: A belief that stuttering is caused by the misdiagnosis of typical disfluencies as stuttering
Capacities and demands: A view of stuttering (often called the demands and capacities model) that suggests that stuttering results when the demands (e.g., pressure to talk rapidly) put on a child’s speech are greater than the child’s capacity for fluency (e.g., capacity to manage the complex components of spoken language production at a high rate)
Communicative failure and anticipatory struggle: A view of stuttering that supposes that stuttering begins when a child experiences problems with communication (e.g., having many repetitions or being told he must try harder to say sounds correctly) and then develops a fear of having difficulty, which then causes tension and fragmentation of speech
Primary stuttering: Early stuttering, near the onset of the disorder, that is characterized by loose, easy repetitions. For those children who begin stuttering in this fashion, it is assumed that at first they are not aware of their stuttering and do not react to it
Secondary stuttering: Stuttering characterized by tension and struggle and sometimes by avoidances. In some views, this type of stuttering is thought to be a reaction to primary stuttering, as the child becomes self-conscious and frustrated by her difficulty with speaking
Myelination: The development of an insulating sheath around nerve fibers, increasing the speed and integrity of neural transmission. Myelination occurs primarily before age 5 and is achieved in left-hemisphere frontotemporal tracts before the right-hemisphere tracts
Sensory-motor modeling: The process of building “maps” that code the motor commands to hit the auditory (and kinesthetic and proprioceptive) targets that the speaker intends; a bidirectional process
Behavioral inhibition system: A view developed by psychologist Jeffrey Gray (1987) that humans are endowed with a hard-wired protective response to frustration and fear; the body’s response in this situation is freezing, flight, or avoidance
Autonomic reactivity: The tendency of the autonomic nervous system to respond quickly and strongly to various stimuli. The sympathetic part of the autonomic nervous system is responsible for the flight-or-fight response and causes the heart, respiratory system, and other organs to be ready for action
Anomalous neural organization: Brain structure and function that differs from the typical. This often (but not always) causes problems because structures not well suited for certain activities (such as speech and language) are used and because subcomponents of speech and language may be located at some distance from each other, and neural transmission of information may not be efficient
What are theories, and what can we learn from them? A theory puts together findings in a systematic way so that past phenomena are explained and future ones are predicted. A theory about tsunamis explains how they are caused by earthquakes on the ocean floor and predicts that when a large undersea earthquake occurs again, another tsunami will occur. A theory about stuttering would take the many facts, findings, and observations that you have been reading about in the first five chapters and put them together to explain why one person stutters and another does not. A complete theory would also explain why a person stutters on some words and not others or in some situations and not others and why stutterers do the things they do when they stutter. When a theory can explain these things well, it can lead to effective treatment. When we know what causes stuttering, we will have a better chance of being able to modify it and perhaps even prevent it.
Scientists often use the word “theory” to mean a formal set of hypotheses that explain the important causal relationships in a phenomenon. These hypotheses are then tested, and the theory may be thrown out, improved, or partially confirmed as a result. The field of stuttering research and treatment hasn’t developed far enough to have a formal theory of stuttering, although there are a number of informal theories that might be called theoretical perspectives or theoretical models.
In this chapter, I present several theories of stuttering as well as my own attempt to integrate research and clinical findings. I have organized the theories by the areas covered in Chapters 2 to 5–constitutional, developmental, and environmental factors. These models change every few years as more data are gathered on stuttering and new information is generated in related areas. Without a doubt, the explanations of stuttering I summarize in this chapter, including my own, will be superseded by others in a few years.

Figure 6.1 An overview of theories about stuttering.
I have chosen several contemporary views of constitutional factors to discuss in the following sections. Although the views differ, they are not mutually exclusive. If linked together, they provide us with some interesting notions about what factors might be inherited or acquired and how that might result in stuttering.
Stuttering as a Disorder of Brain Organization
Many studies of both normal speakers and brain-damaged patients have demonstrated that the left hemisphere is dominant for language in most people. This means that areas in the left hemisphere are specialized for processing language and that the right hemisphere is subservient to the left, playing a minor role in the production and comprehension of language.
One early theory of stuttering suggested that it is caused by lack of hemispheric dominance (the Orton-Travis theory of stuttering referred to in Chapter 2). The theory came about in this way. In an atmosphere of intense scientific curiosity and collaboration among researchers at the University of Iowa in the 1920s, Samuel Orton, a neurologist, and Lee Edward Travis, a psychologist and speech pathologist, observed that many stutterers seemed to have been left-handers whose parents changed them into being right-handed (Travis, 1931). They suspected that this change led to conflicts in the control of speech in which neither hemisphere was fully in charge, creating neuromotor disorganization and mistiming of speech, in turn resulting in stuttering. Travis and Orton’s treatment derived from this theory was simply to switch stutterers back to being left-handed. This approach, as you might guess, turned out to be fruitless. Furthermore, there was never convincing evidence that high numbers of stutterers were originally left-handed. Consequently, the original cerebral dominance theory of stuttering languished for many years. But in the 1960s, evidence began to accumulate that stutterers may not, after all, have normal left-hemisphere dominance for language. In the 1970s and early 1980s, more published studies supported this finding.
In 1985, a new version of the cerebral dominance theory of stuttering was proposed. Two neurologists, Norman Geschwind and Albert Galaburda, proposed that many disorders, including stuttering, dyslexia, and autism, resulted from delays in lefthemisphere growth during fetal development that led subsequently to right-hemisphere dominance for speech and language (Geschwind & Galaburda, 1985). The delay in left-hemisphere growth that resulted in these predominantly male disorders was thought to be caused by a male-related factor. Geschwind and Galaburda hypothesized that these delays might result from fetal exposure to excess testosterone during embryonic development. So far, however, no evidence has been found to support their hypothesis about testosterone. In fact, Neilson, Howie, and Andrews (1987a) provided some evidence against this hypothesis, but the idea of a delay in left-hemisphere development continues to be of great interest.
Geschwind and Galaburda’s theory suggests that a delay in left-hemisphere growth and development may affect speech and language for the following reasons. Various left-hemisphere structures that evolve during embryonic development appear to be especially suited for speech and language functions. As these structures develop, specialized nerve cells that are genetically programmed to sprout the neural connections for speech and language processes disperse from their point of origin in the “neural tube,” where the central nervous system is formed. These nerve cells normally migrate to previously developed structures in the left hemisphere that are appropriate for their specialized functions. But if development of left-hemisphere structures is delayed, cells migrating from the neural tube may not receive the “homing beacon” they need to reach the left hemisphere. Instead, these specialized cells receive signals from the more developed right hemisphere and migrate there instead. These specialized cells then organize themselves as “networks” of neural activity in the right hemisphere for processing of speech and language. However, because the right hemisphere is not designed by its architecture and interconnections for this function, speech and language operate inefficiently there, like the Internet search engine Google trying to access information through the postal service.
Although speculation about atypical cerebral localization in stutterers has received some support from research, especially from the recent brain imaging studies reviewed in Chapters 2 and 3, details of how stuttering behaviors result from abnormal localization are another issue requiring other theories that will be described in a later section of this chapter.
William Webster has proposed another version of the view that stuttering results from anomalous cerebral organization. As I noted in the discussion of studies of nonspeech motor control in Chapters 2 and 3, Webster (1993a) studied the effect of interference on one task (sequential finger tapping) caused by another task (turning a knob in response to an auditory signal) done simultaneously by the other hand. He concluded that people who stutter have normal localization of speech and language in the left hemisphere but that their left-hemisphere structure for speech planning and sequencing, the supplementary motor area (SMA), is especially vulnerable to disruption by activities in other areas of the brain. Webster suspected the SMA because it is strongly connected to the motor cortex as well as to subcortical motor areas and is known to be involved in the initiation, planning, and sequencing of motor activities. The left SMA, located near the corpus callosum, receives input coming across this bridge from the right hemisphere as well as input from the left hemisphere itself. Because of its location and its multiple inputs from both hemispheres, the left SMA might be highly susceptible to disruption by excess activity from either hemisphere. Webster further suggested that individuals who stutter often may have overactive right hemispheres and speculated that overflow of right-hemisphere activation, especially from right hemisphere–regulated emotions (e.g., fear), would disrupt SMA functions in planning, initiating, and sequencing speech motor output.
Stuttering as a Disorder of Timing
Several authors believe that the known facts about stuttering point toward a disorder of timing. For example, Van Riper (1982, p. 415) stated that “when a person stutters on a word, there is a temporal disruption of the simultaneous and successive programming of muscular movements required to produce one of the word’s integrated sounds…” Building on Van Riper’s view, Kent (1984) marshaled several lines of evidence to support a hypothesis that stuttering arises from a deficit in temporal programming. He speculated that this deficit reflects the inappropriate localization of speech and language functions to the right hemisphere that results in an inability to create the precise timing patterns needed to perceive and produce speech efficiently. Like a conductor of a symphony orchestra who determines when each section plays, as well as its speed or tempo, mechanisms in the brain control the rate at which we speak and the order of movements for producing sequential sounds. Just as a conductor integrates the timing of an orchestra’s several sections, the brain must coordinate complex timing relationships for phonemes, syllables, and phrases of speech.
Kent (1984) suggested that the inability to perform precise timing functions consistently may stem from a stutterer’s left hemisphere being less well developed than the right hemisphere (cf., Geschwind & Galaburda, 1985). Because the left hemisphere is specialized for processing brief, rapidly changing events such as those needed for fine motor control of verbal output, a person who stutters may be disadvantaged when trying to process at the speed required for normal speech.1 This central timing function, Kent points out, must not only regulate left-hemisphere aspects of speech production but must also integrate the production of rapid, left hemisphere–generated speech segments with the slower prosodic elements of speech.
Kent also noted that emotion may play an important role in disrupting the timing of the speech of someone who stutters. As I indicated earlier, the right hemisphere is believed to be heavily involved in the regulation of certain negative emotions. The stutterer’s deficit, then, may be that his timing functions for speech are arranged so that they are (1) less efficient than those of nonstutterers and (2) vulnerable to interference by right-hemisphere activity during increased emotion. How this deficit causes the repetitions, prolongations, and blocks we hear in stutterers’ speech is not explained in this theory.
Stuttering as Reduced Capacity for Internal Modeling
Another view of constitutional factors in stuttering was advanced by Megan and Peter Neilson, whose research on stutterers’ tracking abilities was reviewed in Chapter 2. The Neilsons proposed that the repetitions of beginning stutterers are the result of a deficit in their ability to create and use “inverse internal models of the speech production system” (Neilson & Neilson, 1987). This rather complicated sounding model can be easily understood if we go back to an assumption about how children learn to speak.
During the first year of life, infants store up perceptions of the speech sounds they hear around them and begin to play with speech sounds, trying to imitate what they hear. Gradually, as they grow older, children learn how to make these sounds accurately. Some scientists, like the Neilsons, believe that too much of the brain’s neural resources would be required if children had to remember each of the movements needed to produce each sound of their language in every possible phonetic context. Instead, children are thought to develop a mental “model” of the relationship between their speech movements and sounds they hear. Just as someone beginning to play a trombone must learn the relationship between the movements of the trombone “slide” and the sounds that result, experienced trombonists have established mental models of the relationship of their arm movements to the sounds produced and are able to move the slide to produce a desired sound without having to think about it in any deliberate way.
A child, then, develops a mental model of the relationship between speech sounds and motor commands. The mental model in the brain might be called a sensory-motor model for speech, which the Neilsons call an “inverse internal model” of how speech is produced. It is an “inverse” model because it transforms or inverts sensory targets (i.e., heard speech sounds) into the motor commands needed to produce them. As infants learn to produce the sounds they hear, they constantly use and refine their sensory-motor model for speech. They plan a word or sentence in terms of what it should sound like (the target) and then rely on their sensory-motor model to generate the motor movement commands that will produce the speech targets they are trying to hit.
The process of learning to speak is something like learning to drive a car. At first, keeping the car on the road requires constant vigilance. But as we learn the relationships between turning the wheel, stepping on the accelerator, and going where we want, the linkage becomes automatic, even when driving a stick-shift vehicle in stop-and-go traffic. Moreover, the linkage is refined as we encounter different driving conditions and different cars (e.g., cars with loose steering wheels and sticky accelerators). Just as drivers establish sensory-motor models for driving, children develop sensory-motor models for speaking.
Figure 6.2 is a schematic depiction of how the brain may transform desired sensory (perceptual) targets into motor commands for speech. In the figure, the desired output (the word or phrase, e.g., that a child intends a listener to hear) is fed into the internal inverse model of the speech production system. Here, the desired output is entered as sensory code of its expected auditory and kinesthetic results, which is “inverted” by the model to generate its output as movement codes or motor commands. Experience, practice, and vocal play help the child to acquire these inversions or transformations. Moreover, this internal model is continually updated as a child’s speech and language skills mature and the speech production system changes with age. The internal model’s motor commands are sent to the muscles of the speech production system, whose coordinated contractions produce the acoustic output that result in a planned utterance. Concurrently, ongoing planning and feedback of this process are fed into the modeling circuitry.

Figure 6.2 Schematic of the inverse internal model theory of speech production.
Let us retrace our steps for a moment; when motor commands are sent to muscles, a copy of these commands, which is called the “efference copy” by motor physiologists, is also sent to the modeling circuitry. Here, efference copy is transformed into its hypothetical output, which is a model, or template, of the output that should be produced based on the motor commands. This hypothetical output is continuously compared with feedback on the current positions and movements of the speech mechanism so that the inverse internal model can update its ongoing motor commands, if necessary, to produce the desired output more accurately. These components of the speech production process are assumed to involve the corticocerebellar structures and pathways that are commonly described in neural models of speech output (e.g., Neilson & Neilson, 1987; Neilson, Neilson, & O’Dwyer, 1992; Neilson & Neilson, 2005a, 2005b).
The Neilsons and their coworkers have used the inverse internal model of the speech production system to understand the performance of stutterers in experiments that tested their ability to track an auditory tone that changed unpredictably (Neilson, Quinn, & Neilson, 1976). The subjects heard an unpredictably changing “target” tone in one ear and a “cursor” tone, which they could control with a handheld device, in the other ear. Their task was to track the pitch of the target tone with the cursor tone as accurately as possible. The Neilsons’ experiments found that stutterers were poorer than nonstutterers in tracking auditory tones that went up and down in pitch. Stutterers were still poorer than nonstutterers even after practicing the task. These findings suggested to the Neilsons that if stutterers had difficulty learning the relationships between the sounds they want to say and the movements required to produce them as young children, they would also, therefore, have difficulty making the sensory-tomotor and motor-to-sensory transformations required by the tracking tasks.
However, this difficulty would not always result in stuttering. When circumstances don’t call for much of the brain’s functional capacities in speech and language areas, stutterers should be able to compensate for their slight weaknesses. On the other hand, when large portions of the brain’s functional capacity are allocated for language tasks, such as choosing new or unfamiliar words or constructing complex sentences, the diminished neural capacity cannot be accommodated, and more repetitions would result. As these researchers put it, “whether one will become a stutterer depends on one’s neurological capacity for these sensory-to-motor and motor-to-sensory transformations and the demands posed by the speech act” (Andrews et al., 1983).
How do these intermittent deficits in available functional neural capacity result in the symptoms of stuttering? This theory attempts to account only for the core behaviors of early stuttering, that is, repetitions and prolongations. According to the theory, repetitions and prolongations result from inadequate transformations of sensory targets, transformations that should generate the motor commands for speech. A speaker with reduced functional neural capacity may begin to speak but be unable to plan and carry out the rest of his utterance without disruption. Repetitions or prolongations may occur if a speaker is attempting to push ahead with speech while his brain is still planning the syllables that follow and how to link them with the initial sound.
Recently, other researchers have also used the concept of inverse internal models to explain the behaviors of stuttering. The Neilsons’ view was echoed in the perspective described in Guenther (1994), which was later adapted by Max, Guenther, Gracco, Ghosh, and Wallace (2004) to propose a theoretical model of stuttering based on unstable or insufficiently activated internal models. One of their hypotheses parallels the Neilsons’ proposition that some children are predisposed to stutter because of the difficulty learning the relationships between their motor commands and the desired acoustic output. This difficulty would result in an inaccurate inverse internal model of the speech production system (Fig. 6.2), which would generate output that would not match the desired perceptual consequences. The speech production system would then “reset” itself to try again, producing repetitions (one for each mismatch). This resetting process would continue until the child’s error-correction process could update the model sufficiently to make the output match the consequences. If this could be done quickly, only one or two repetitions would occur; if not, many repetitions would occur. The Max team’s proposal has other hypotheses and an extensive review of the literature to support them.
Their proposal is particularly effective in relating various stuttering phenomena to aspects of the model. For example, the findings that stutterers’ movements are slower during fluent speech (see Chapters 2 and 3) and the evidence that slow speech can induce fluency are both explained by the possibility that a slower rate of speech production would allow the stutterer more time for feedback to update the internal model. With a properly updated internal model, the actual speech acoustic output would match the intended perceptual consequence. Therefore, the system would not produce the repetitions that are thought to be a result of inaccurate speech output that doesn’t match the intended perceptual consequences. In other words, slower speech makes corrections possible while a syllable is being produced rather than after it is completed. Note that if this hypothesis is accurate, errors would be found in the stutterer’s unsuccessful repetitions.
Recent work by Cykowski, Fox, Ingham, Ingham, & Robin (2010) has updated this perspective on stuttering as a deficit in using inverse internal models for speech. Specific pathways in the brain used for sensory-motor integration–within the superior longitudinal fasciculus (SLF III)–have been found to be less efficient in adults who stutter compared to adults who do not. This new information will be described in detail when we present an integrated model of stuttering in the last section of this chapter.
Stuttering as a Language Production Deficit
Many researchers have been intrigued by the influence of linguistic factors on stuttering. For example, stuttering often begins when a child enters a period of intense language development. Similarly, stuttering is most frequent when the load on language functions is heaviest (e.g., in longer utterances, at the beginnings of sentences, and on longer, less familiar words) (Bloodstein, 2002). These factors have prompted several theorists to propose that stuttering reflects an impairment in some aspect of spoken language. I use the term “spoken language” because these theorists believe the major problem is not in the motor execution of speech, but rather in the planning and assembly of language units, such as phonemes, that occur before speech is produced.
Herman Kolk and Albert Postma (1997) developed the “covert repair” hypothesis to explain stuttering from a language production point of view. They believe that both stuttering and normal disfluencies result from an internal monitoring process that we all use to check whether what we are about to articulate is exactly what we mean to say. Perhaps this may be clearer if we imagine for a moment that language production is like a factory making bicycles (Fig. 6.3). The factory must monitor the quality of its bicycles by checking them at different stages. Some quality control checks occur after the bicycles leave the factory, when factory workers themselves ride the bicycles and tell the factory about any defects they find. In speech and language production, this is like a speaker’s auditory feedback (the sound of your own words as you are speaking them).

Figure 6.3 Quality control in a bicycle factory as an analogy for part of the language production system in the brain.
The bicycle factory might also use another quality control process, one that occurs inside the factory before the bicycles are shipped out. This is like our internal monitoring process of speech and language. Without being aware of it, we check the “phonetic plan” for what we are about to say before we articulate it. This allows us to detect potential semantic, syntactic, lexical, and phonological errors before they are produced. Just as the production line in a bicycle factory would have to be halted when a defect is detected, speech production is interrupted when our internal monitor detects an error in our phonetic plans. Repairs need to be made before production can continue. Kolk and Postma (1997) believe that the halting of production and the repair process cause the disfluencies of both normal speakers and individuals who stutter.
To Kolk and Postma (1997), the most common stuttering disfluencies (repetitions, prolongations, and blocks) are the result of correcting or “repairing” the phonological (rather than semantic, syntactic, or lexical) errors detected in the phonetic plan before they are spoken. In the case of part-word repetitions, if a speaker detects an error in the final part of a syllable (e.g., the /p/ in “cup”), he restarts the phonological encoding process ( “cu-cu-”) and keeps going until the phoneme is encoded correctly and the entire syllable can be produced. In contrast, prolongations are thought to occur when the phoneme of a word or syllable preceding the error is a continuant (e.g., the /l/ in the word “lip” when the error involves the vowel). In this case, the continuant, /l/, is prolonged until the speaker successfully encodes the vowel, /i/, following /l/.
Blocks are thought to result from errors in the initial sounds of words or syllables. When an error is detected, speech production is halted for repairs, but the speaker may try to plunge ahead, building up muscle tension, unaware of the automatic error detection and repair that is in progress. Kolk and Postma (1997) suggest that stutterers are prone to have more phonological encoding errors because they are constitutionally slower in encoding and need more time than a typical conversational rate gives them. In various articles, Kolk and Postma lay out the evidence supporting their views and suggest, among other things, that the benefits of a slower speech rate on stuttering are derived from the greater amount of time that stutterers have for phonological encoding.
Several years before Kolk and Postma (1997) published their covert repair hypothesis, an innovative language production view of stuttering was published by Wingate in The Structure of Stuttering: A Psycholinguistic Approach (1988). In this book, he reviewed linguistic and neurological research on stuttering and hypothesized that stuttering results from a dyssynchrony of functions in the left and right hemispheres, as well as subcortical structures. These different areas, Wingate suggested, are responsible for different components of language planning and production, such as consonants, vowels, and prosody. He theorized that when speakers produce the initial portion of a syllable, the consonant, vowel, and prosody must be synchronously blended. If some component lags behind at this critical moment, the result is a disruption in speech production that we observe as stuttering. Returning to our imaginary bicycle factory, it is as if the wheels, gears, and frame all must be assembled at the same time on a high-speed assembly line. If one component is delayed, production is halted. However, Wingate does not explain how this halt appears in speech as a repetition, prolongation, or block.
Perkins, Kent, and Curlee (1991) proposed another theory of stuttering as a deficit in language production. These authors suggested that stuttering results from a dyssynchrony between two components of language production. The “paralinguistic” component is a right hemisphere–controlled social-emotional process that is responsible for vocal tone and prosodic functions. The other component is linguistic and involves a left hemisphere segmental system that is responsible for the content and structure of language (semantics, syntax, and phonology). The two components must be integrated before spoken language is produced. If one lags behind the other for whatever reason, the resulting dyssynchrony produces disfluency.
Perkins, Kent, and Curlee (1991) add two elements to this dyssynchrony that must also be present if the resulting disfluency is stuttering, rather than just a normal disfluency. First, the speaker must experience time pressure either from an outside source or an inner feeling, so that he continues trying to speak even though the dyssynchrony in paralinguistic or linguistic processes has resulted in an incomplete or anomalous speech motor program. Second, the speaker must experience a feeling of “loss of control,” which arises from being unaware of why he cannot say the word.
In our imaginary bicycle factory, Perkins, Kent, and Curlee’s theory might be characterized as a production line that stops automatically whenever one of the two major subcomponents of a bicycle is not ready for assembly. However, the boss in this factory demands that the production line move rapidly, and the workers panic if the production line grinds to a halt. They frantically keep trying to restart production even though they don’t know what the problem is or how to fix it.
Stuttering as a Multifactorial Dynamic Disorder
For almost 25 years, Anne Smith has been carrying out a systematic program of research on stuttering, developing a theory that at the core of stuttering is a motor speech disorder, the appearance and severity of which are influenced by a multitude of cognitive, linguistic, and psychosocial factors (e.g., Smith, 1999; Smith & Goffman, 2004; Smith & Kelly, 1997; Zimmerman, Smith, & Hanley, 1981). In arguing for the multifactorial nature of stuttering, Smith quotes Van Riper (1982) who makes the point that not only are there multiple factors acting in concert that determine if individuals stutter but that different individuals will have unique combinations of factors–different amounts of various factors–that determine their own stuttering fate.
Portraying stuttering as a multifactorial dynamic disorder, Smith thinks it is inappropriate to search for a single underlying “cause” of stuttering but instead thinks it’s important to look for which factors interact in stuttering and determine how they interact. A good example of the way this view is manifest in Smith’s research is the finding that when a group of individuals who stutter produce utterances that are longer and more linguistically complex, their speech motor coordination becomes more variable (movements of articulators are less regular when standard deviations are computed for many repetitions of a phrase), compared to a group of nonstutterers (Smith, Sadagopan, Walsh, & Weber-Fox, 2010). Despite the greater variability in their coordination, the individuals in the stuttering group did not overtly stutter as they produced the utterances in this experiment. Exactly how this greater variability does result in stuttering from time to time will probably await further study of other factors that can be manipulated in these experiments, such as psychosocial stress.
The multifactorial, dynamic view characterizes stuttering as a “dynamic” disorder because the “stuttering events” of repetitions, prolongations, and blocks are seen as only the outward manifestation of an underlying, ever-changing process resulting in these events erupting to the surface from time to time.2 So in the experiment described above, there may be increasing speech motor instability as linguistic load increases, and if psychosocial stress were increased (as part of the underlying, ever-changing process), some stuttering events might surface. Perhaps this would occur in some individuals but not others.
Smith and her colleagues’ research on speech motor instability (variability of movements) hints at a substrate that might be related to early or primary stuttering, but what can explain the secondary reactions of tension and struggle? What causes stuttering to change from brief, easy repetitions or prolongations of words and syllables that are hardly noticed by the child or listeners to the long, tense blockages that frustrate, embarrass, and upset both the speaker and often his audience?
Smith and her colleagues have done some important research on this aspect of stuttering as well. Some of this work has indicated that some moments of stuttering are characterized by rapid (5 to 12 Hz), rhythmic, oscillatory neural input to the muscles of speech so that they contract in a rapid, tremor-like way (Smith, 1989). Not every stutterer’s muscles show these tremors during stuttering, however, and they may not appear in younger children who stutter but only in older children whose stuttering has persisted for some time. In one study, Kelly, Smith, and Goffman (1995) found that these neural oscillations were present in the stutters of the three older children in their study (ages 10 to 14 years) but were absent in the stutters of the seven younger children (ages 2 to 7 years). The researchers noted that such tremors may appear in the stuttering of only those individuals who have stuttered for some time and have developed maladaptive reactions. As Kelly, Smith, and Goffman (1995) pointed out, it is also possible that these tremors are evoked or magnified by autonomic arousal or the emotion that arises in response to the expectation or occurrence of speech difficulties.
Several writers have linked stress with stuttering and suggested that the tiny tremors that appear in everyone’s muscles may be amplified by emotion to a level that interferes with talking (Fibiger, 1971, 1972; Van Riper, 1982; Weber & Smith, 1990). The effects of emotion on tremor may provide a physiological explanation of how the mild disfluencies of young children become the more severe blockages we see as beginning stuttering evolves into advanced stuttering. Even mild disfluencies may trigger emotional responses in some children that result in increased tremors that block speech. Emotional responses may also explain the unusual cases of severe blocks at the onset of stuttering, especially when stuttering begins during conditions of stress and strong emotion. The interaction of very strong emotion with a child’s vulnerable speech motor system may create sudden severe stuttering because it amplifies tremors in the speech musculature. Such tremors may be analogous to the quivering lip of a toddler who is about to burst into tears when frightened by a barking dog or by a screaming parent. Just how magnified tremors block or slow the forward movement of speech is not known, but Van Riper’s (1982, p. 126) description of tension and tremor may give us some clues. He suggested:
What usually seems to happen is that tremors begin when the stutterer creates a fixed closure, invests its antagonistic musculatures with tension, and then suddenly produces an increase of air pressure behind or below the closure. At the moment this increase occurs, the antagonistic musculatures become suffused with a sudden burst of further tension and the stuttering tremor comes into being. Then it persists…
Why stutterers would create fixed closures is a mystery. Perhaps it is one of the body’s responses to fear that Gray (1987) described as a behavioral inhibition system, which I will discuss later in the section on secondary stuttering and temperament.
The three views presented in this section represent three different conceptualizations of how developmental or environmental stresses (or both) result in stuttering. One view is called the “diagnosogenic” view because it proposes that stuttering begins when parents mistakenly diagnose normal disfluency as stuttering. The other two views look more broadly at circumstances under which stuttering might arise. For example, the communicative failure and anticipatory struggle view assumes that some form of communication difficulty precipitates stuttering, whereas the capacities and demands view presumes that almost any developmental or environmental pressure may precipitate it. As you read this section, keep in mind that these three views differ not only in their concept of the roles that development and the environment play but also in their specificity. The first view (i.e., diagnosogenic) proposes that specific factors create stuttering; the last view (i.e., capacities and demands) describes how many different variables may interact to produce stuttering. The specificity of the remaining view (i.e., communicative failure) lies somewhere in between the two others.
Diagnosogenic Theory
In the 1930s, Wendell Johnson and other researchers at the University of Iowa began studying the onset of stuttering in children. As Johnson examined the speech of young stutterers and nonstutterers, he noticed a similarity. The most common disfluencies of both groups were repetitions. As Johnson contemplated this observation, he was struck by the possibility that all of these children may have had the same disfluencies to begin with but that those who became stutterers developed more serious disfluencies by overreacting to their repetitions. But why? Johnson speculated that perhaps their parents or other listeners mislabeled their repetitions as stuttering and in so doing, made the children so self-conscious that they tried to speak without any disfluency. Their efforts to avoid all disfluencies may have become what is generally regarded as stuttering (Johnson et al., 1942).
Johnson’s hypothesis came to be called the diagnosogenic theory, meaning that stuttering was caused by its diagnosis or in this case, misdiagnosis. It was the most widely accepted explanation of stuttering throughout the 1940s and 1950s. It pinpointed environmental factors as the sole cause of stuttering by placing the blame on the negative reactions of parents and other listeners.
Johnson and his associates continued gathering data on the disfluencies of stuttering children and their nonstuttering peers to further support the diagnosogenic theory. The results of several studies were summarized in a landmark book, The Onset of Stuttering (Johnson & associates, 1959). Table 6.1, taken from Johnson’s book, gives an overview of the similarities and differences in the disfluencies reported by stutterers’ and nonstutterers’ parents. Johnson interpreted these data as showing similarity between the disfluencies of stuttering and nonstuttering children; both groups of children were reported to show at least some of each type of disfluency. He suggested that the same disfluency types that parents of nonstuttering children considered normal were reported by parents of stuttering children as the earliest signs of stuttering. Johnson used this as evidence to support the diagnosogenic hypothesis that the problem was parents’ interpretation of their child’s disfluencies, or as some often put it, “the problem was not in the child’s mouth but in the parent’s ear.”
Table 6.1 Percentage of Parents of Stutterers and Nonstutterers Who
Reported Child Was Performing Each Speech Behavior When They First Thought Child Was Stuttering

From Johnson, W., et al. (1959). The onset of stuttering. Minneapolis: University of Minnesota Press. Copyright © 1959 by the University of Minnesota. © 1987 Edna Johnson. Reprinted with permission of the University of Minnesota Press.
It should be noted that later authors interpreted the data in Table 6.1 quite differently than Johnson did. The data were seen as evidence that the two groups of children were different at the onset of stuttering. McDearmon (1968), for example, argued that these findings showed that the disfluencies of normal children were notably different from those in the stuttering children. Syllable repetitions, sound prolongations, and complete blocks were reported to have occurred much more frequently in the stuttering children, whereas phrase repetitions, pauses, and interjections were reported more frequently in the nonstuttering control group. This reinterpretation of Johnson’s evidence and new findings about genetic and constitutional factors in stuttering have caused the diagnosogenic view of stuttering to be largely abandoned.
To illustrate the diagnosogenic view, I take an example from a master’s thesis that Johnson directed (Tudor, 1939). At that time, the diagnosogenic theory had not been formally proposed, but undoubtedly Johnson and others must have entertained the possibility that labeling a child as a stutterer would create more hesitancy in his speech. This thesis was an exploration of that idea. Johnson’s student, Mary Tudor, screened all the children at a nearby orphanage for speech and language disorders. Tudor selected six children who were normal speakers, but she told these children that they should speak more carefully because they were making errors when they talked. She warned them that they were showing signs of stuttering. She also cautioned caregivers that these children should be watched closely for speech errors and corrected when they slipped up. After several months, Tudor went back to the orphanage and found that a number of these children showed stuttering-like behaviors. Although she tried to treat them, at least one child was reported to have continued stuttering for some time thereafter (Silverman, 1988). Tudor was remorseful about these results and regretted conducting this experiment (Zebrowski, personal communication). Nonetheless, it reinforced Johnson’s strong conviction, which he held throughout his career, that if a child is made self-conscious about his normal disfluencies, he may begin to stutter.
Communicative Failure and Anticipatory Struggle
The theoretical view of communicative failure and anticipatory struggle, developed by Oliver Bloodstein (1987, 1997; Bloodstein & Ratner, 2008), proposes that stuttering emerges from a child’s experiences of frustration and failure when trying to talk. The child’s original difficulty in talking may be the typical disfluencies of childhood, but other frustrations might instead be the provocation for stuttering. Many types of communication failure may lead the child to anticipate future difficulty with speech and thus increase tension. It is common, Bloodstein noted, to find delays in the development of articulation and language, cluttering, and other speech problems in the histories of children who begin to stutter. Table 6.2 lists some of the circumstances that Bloodstein suggested might cause some children to believe that speaking is difficult. If a child cannot make himself understood or is penalized for the way he talks, he may begin to tense his speech muscles and fragment his speech, reactions that become the core behaviors of the child’s stuttering. And these behaviors in turn result in more frustration and failure in communication, which the child anticipates with dread.
Table 6.2 Experiences That May Make Some Children Believe Speaking is Difficult

Other aspects of the child’s “internal” and “external” environments and his development also play important parts. The child’s personality may be perfectionistic, or he may harbor needs to live up to parental expectations. His family may have high standards for speech, find any speech abnormality unacceptable, or otherwise pressure the child to conform to standards beyond his reach. The presence or absence of these sorts of developmental and environmental pressures may cause some children to interpret an articulation difficulty, language problem, or disfluency as a failure, whereas other children only shrug it off.
This perspective on stuttering accounts for the wide variability of disfluency among children. Most normal children experience temporary frustration when learning to talk as they produce the mild fragmentations of speech we associate with normal disfluency. Children who stutter for just a few weeks may encounter unusual difficulty when first learning to talk but soon master the fundamentals and feel successful. Children who become chronic stutterers may be those who repeatedly experience communication failure and grow up in an environment fraught with communicative pressure.
Here is a case that illustrates some of the environmental pressures that some children who begin to stutter may experience. Susan grew up in the oil fields of Oklahoma, where her parents set themselves apart from the rest of the community by their aloof manner and precision of speech. They raised their children to feel that they were more cultured than their neighbors; in fact, Susan’s father would often say, “We speak better than other people.” Unfortunately, Susan’s speech development was delayed. When she did begin talking in sentences at about age 3, she began to stutter with mild repetitions. When she started school, she worried that her father was embarrassed by her speech. Then she tried to speak better and began to push out the words instead of repeating the first parts of them. She soon developed severe secondary stuttering.
Although we have no way of knowing for sure, Susan’s critical father may have been a major factor in the onset of her stuttering. However, many children grow up in families that are critical of speech but don’t develop stuttering. Perhaps both a constitutional deficit, which led to her delayed speech development, and family pressure for perfect speech were necessary to produce Susan’s stuttering. Neither may have been sufficient by itself to create stuttering, but together they may have been enough to tip the balance.
Capacities and Demands
A third interactional view of stuttering onset is proposed by the capacities and demands theory. Others have called this a “demands and capacities” view, but I prefer to put capacities first, because they exist in children before demands are placed on them. This view suggests that disfluencies as well as real stuttering emerge when a child’s capacities for fluency are not equal to speech performance demands. Earlier in this chapter, I briefly discussed a narrow version of this view in describing the reduced capacity for internal modeling theory of stuttering. Andrews and colleagues (1983) stated that “whether one will become a stutterer depends on one’s neurological capacity … and the demand posed by the speech act” (p. 239). These authors indicated that some demands come from the rapid development of language between ages 3 and 7 years. Other demands may come from fast-talking parents, whose speech rates may be hard for a child to keep up with. Demands for speech performance sometimes come from within the child, sometimes from outside stimuli, and sometimes from both.
Joseph Sheehan (1970, 1975) expressed an early variation of the capacities and demands view when he wrote that, “a child who has begun to stutter is probably a child who has had too many demands placed on him while receiving too little support” (Sheehan, 1975, p. 175). The demands that Sheehan pinpointed were primarily those of parents who have high standards and high expectations for their child’s behavior. The support he refers to appears to be the environment’s capacity to provide love, care, and encouragement. In addition, he believed that, “there are persisting reasons for retaining the possibility that some kind of physiological predisposition for stuttering exists” (Sheehan, 1975, p. 144). Thus, Sheehan, who is best known for a theory that stuttering is learned, professed the view that stuttering is precipitated by the demands of the environment interacting with a predisposition to stutter.
Starkweather (1987) added considerable detail to the concept of capacities and demands as an explanation of stuttering onset and development. The normal child’s capacities, he points out, include the potential for rapid movement of speech structures in well-planned sequences that are coordinated with the rhythms of his language. Demands on the child include those of her internal environment, such as her increasingly complex thoughts to be expressed, which require more sophisticated phonology, syntax, semantics, and pragmatic skills. The external environment often places demands on the child’s fluency through parents’ interactions. Parents may ask questions rapidly, interrupt frequently, and use complex sentences choked with big words. They may show impatience about the child’s normal disfluencies and may make the child feel that she meets their expectations only when she performs at high levels. These kinds of interactions can stress any child but are likely to push a slowly developing child to try to speak beyond her capacity for fluency.
Because a child’s capacities develop in spurts and environmental demands fluctuate, stuttering may wax and wane in rapid cycles. A child may be highly fluent for a day or a week when he has mastered new speech and language skills and when external demands are low. But his stuttering may suddenly flare up if his capacities become strained by his efforts to use more advanced syntax or if the demands of the external environment suddenly increase when his fast-talking, interrupting, big-city cousins arrive for the Fourth of July holiday weekend.
The capacities and demands view provides a way to account not only for the day-today variability of stuttering within an individual but also the great differences between one individual who stutters and another. As Adams (1990) pointed out, some children may grow up in an environment with normal levels of demand but have limited speech production capacities. Others may have normal capacities for speech production but grow up with excessive demands for rapid, fluent speech.
Treatment based on this model would begin with a careful evaluation of the child’s capacities and the demands in her environment. Therapy would be designed to enhance capacities, decrease demands, and provide support for the child and her family while these changes are taking place. Starkweather and his colleagues have used this approach to formulate a sensible and effective program of stuttering prevention (Gottwald, 2010; Gottwald & Starkweather, 1984, 1985; Starkweather & Gottwald, 1990; Starkweather, Gottwald, & Halfond, 1990). Figure 6.4 depicts the ratios of capacities and demands in a child predisposed to stutter. In one view, the demands are greater than the child’s capacities and stuttering occurs. In the second, the demands are lessened, and although capacities stay the same, stuttering is diminished.

Figure 6.4 Two different ratios of capacities and demands and their hypothesized effects on fluency.
To illustrate the capacities and demands view more fully, the following case is from my own experience. Gina was a bright, happy 7-year-old. Her mother had been a severe stutterer as a child, but through treatment and her own perseverance, she had largely recovered. When Gina began the second grade, she had no history of stuttering or any problem with school. Some time before Christmas that year, however, when her class was learning to read, Gina began to dislike school, and her mother soon discovered that she was having problems academically. After some testing, it was discovered that she had a learning disability that had not been apparent before; however, once reading was required, it became obvious. As Gina struggled to cope with her reading problem throughout the rest of the second grade, she began to stutter. Over the course of the next two years, she stuttered noticeably but did not receive therapy. She was, however, given extra help for her reading disability. By the fourth grade, Gina was making headway with reading, and her stuttering had diminished to an inconsequential level without treatment.
Although there are various ways to account for the onset of Gina’s stuttering and recovery, a capacities and demands view would see it this way: Gina was predisposed to stutter, but it lay dormant until she was faced with the challenge of reading. Reading, at least when first learned, involves a highly conscious use of linguistic processes, in contrast to the more automatic linguistic processing used in listening and speaking. Consequently, learning to read puts a heavy demand on the pool of available resources that are also used for speech and language processing. Such demands may result in a reduced capacity (i.e., fewer available resources) for speech production, which may result in disfluency for a vulnerable child. In this case, Gina did not seem to develop a persistent fear of speaking as a result of her stuttering. Thus, when she overcame her initial reading difficulty and reading became more automatic (i.e., demanded fewer resources), her available capacity for speech processes increased, and she “outgrew” her stuttering.
Once again, the reader is reminded that the capacities and demands view is a model for describing relationships that appear again and again but are not well understood. As such, its major function is to help students and clinicians organize the complex interrelationships of variables associated with stuttering into a set of principles that may guide its treatment and suggest hypotheses for research.
In this section, I will try to draw together some of the theoretical views just described coupled with my own speculations to provide a description of the etiology of stuttering that may guide your assessment and treatment. Figure 6.4 depicts the major components of this perspective.
A Two-Stage Model of Stuttering
Many years ago, a child psychiatrist who stuttered, Charles Bluemel, observed that stuttering begins in most children as repetitions, of which they are hardly aware and to which they don’t react. He thought that over time, many of these children become aware of their disfluencies and react to them by increasing the tension and tempo of their repetitions. These repetitions then become fast, irregular, and halting as children are bothered by them and do what they can to stop them. As they tense further, the repetitions become blocks and sometimes prolongations. This can happen overnight in some children and over a period of months for others. Bluemel (1932) called the beginning behaviors “primary” stuttering and the later reactions “secondary” stuttering. Recently, Bloodstein (2001) concluded that this division of stuttering into primary and secondary phases is an appropriate description of the early development of stuttering. This view of stuttering as having two separate stages or components seems useful to me also and suggests the possibility that each component may have a different etiology. In fact, I think each may have a constitutional basis.
There are reasons to be cautious about embracing this apparently simple view. For example, primary and secondary stages of stuttering may overlap in children because the forces that create these stages wax and wane and make it hard to place them in one phase or another. Moreover, there is evidence that some children begin stuttering in the secondary phase (Yairi & Ambrose, 2005). Despite these exceptions, I will lay out an integrated view of stuttering on the presumption that there are, for most children, two stages of stuttering, and I will describe how the exceptions themselves can be explained in this view.
A Perspective on Primary Stuttering
In Chapters 2 and 3, we reviewed studies of brain structures and functions, sensorymotor coordination, central auditory processing, and hemispheric dominance. These studies support the view that individuals who stutter have differences in the way their brains process sensory information and produce motor output. Many of the brain imaging studies point to structural and functional anomalies in the language- and speech-generating areas of the left hemisphere with possible compensatory activity in homologous areas of the right.
The etiology of these differences in people who stutter appears to be either the result of inheritance or early brain damage. Either of these factors would affect how the brain grows during embryonic development or how the brain responds to injury. Let’s look more closely at this process.
The development of speech and language networks in the brain begins soon after conception with the proliferation, migration, and differentiation of neural cells, a process guided by genetic predisposition and affected by external events, such as experience, injury, and disease (Chase, 1996). As neural cells continue to proliferate and differentiate, millions of synapses are formed, and pathways of communication emerge when clusters of cells send information back and forth in response to stimulation. Cells that communicate readily among themselves become selforganizing, functional neural circuits that perform various tasks. For example, after birth as an infant interacts with the outside world, groups of circuits in the infant’s brain bind together to form “maps” or representations of the outside world to help them process incoming sensory information and produce appropriate motor responses (Edelman, 1992).
If there are anomalies in speech and language areas of the left hemisphere due to inheritance or injury, the developing brain will deal with them in a number of different ways. The most common is by extensive anatomical reorganization, including growth of new fibers, new synapses, and entire new cortical tracts (Hadders-Algra & Forssberg, 2002). This reorganization would attempt to establish the functional circuits necessary for the development of spoken language in whatever structures are available. If reorganization involves relocation of these circuits to areas that have not naturally evolved to serve these circuits, or if reorganization entails neuronal groups being placed at some distance from each other, these circuits will be both inefficient and vulnerable to disruption by other brain activities occurring in nearby areas. Vulnerability to disruption may occur if new circuits or even the original ones are insufficiently myelinated (insulated by a protective sheath), an outcome that would result in less-efficient transmission of neural signals. In fact, several researchers (Karlin, 1947; Cykowski et al., 2010) have suggested that inadequate myelination itself may be the reason why neural circuits for speech in language are inefficient or vulnerable to disruption in the brains of those who stutter.

Figure 6.5 Integrated perspective on stuttering.
Which neural circuits underlying spoken language may be inefficient or vulnerable to disruption? One of the functions that often seems to be atypical in stuttering is sensorimotor processing, particularly auditory-motor processing. Because auditory processing plays a major role in infants’ use of the sounds of adult speech and the sounds of their own babbling, a dysfunction in this area would obviously have an influence on the development of interacting neuronal circuits for speech and language production. The sounds of adult speech give infants auditory targets to aim for when they are learning to speak. The sensorimotor activity in babbling helps a child develop internal models that specify what articulatory gestures are needed to produce desired auditory targets (Guenther, Ghosh, & Tourville, 2006; Hickok & Poeppel, 2007; Neilson & Neilson, 1987, 2005a, 2005b). Moreover, the auditory information from babbling allows the child to adapt his internal auditory-articulatory model to his rapidly growing speech production mechanism (Callan, Kent, Guenther, & Vorperian, 2000). Because these circuits are self-organizing, they may develop a variety of solutions to the auditory processing problem. Some individuals may use homologous right-hemisphere structures for auditory processing, others may continue to use inefficient areas of the left hemisphere, and still others may try to do both.
Four of the brain imaging studies described in Chapters 2 and 3 suggest problems in the very pathways that would be expected to support sensory-motor modeling for speech. The findings of Sommer, Koch, Paulus, Weiller, and Buchel (2002), Chang, Erickson, Ambrose, Hasegawa-Johnson, and Ludlow (2008), Watkins, Smith, Davis, and Howell (2008), and Cykowski and colleagues (2010) suggest that individuals who stutter have less dense bidirectional fiber tracts between sensory and motor areas. If these fiber tracts are less dense, they are probably less efficient for rapid transmission of signals. Let’s step through the process of speech production to see why this matters. Remember that according to the inverse internal model theory of speech production, when individuals plan to generate a word or phrase, they use the inverse internal model system (see Fig. 6.2) to generate motor commands (including making an efference copy–a copy used to rapidly produce a hypothetical output of the motor commands for error-correction purposes) based on the sensory target they are trying to produce. However, this inverse internal modeling system depends on continuous backand-forth communication between sensory and motor areas. Sensory areas supply information about the current tension in muscles and positions of speech structures, while motor areas use this information to plan and carry out the motor commands to speech system muscles to produce speech output. Then sensory areas analyze both the efference copy of the motor plans as well as the actual speech output to correct errors and update the internal model.
Rapid information flow between sensory and motor areas is critical for accurate and fluent speech. If there is a delay in the information needed to generate the motor plans and execute them, repetitions may occur. This often happens after the speaker produces the first sound, syllable, or word. This sound or syllable of the utterance is usually fluent because it can be based on already obtained sensory information about the resting state of the speech system structures. But new information is needed to go forward–information required for production of the next sound, syllable, or word. However, that information is often delayed because it depends on rapid updating of information about the new state of the speech system (sensory analysis) and the rapid analysis of the sound just produced (comparison of efference copy of motor commands and feedback of actual output). Again, see Figure 6.2 for a description of the process.
This description of dyssynchrony in the assembly of components of speech and language production is intended only as a possible explanation of primary stuttering, which is a stage of stuttering characterized by relatively relaxed repetitions and occasional prolongations that typically occur, as Bloodstein (2001, 2002) has suggested, at the beginnings of phrases or sentences. As indicated in the paragraph above, the first sound or syllable may be fluent but the second is often a repeat of the first, li-like this. Most children who begin to stutter outgrow their primary disfluencies as their speech and language systems mature or as they develop effective ways to work around the problem. Other children, however, react to their primary stuttering by increasing the speed and tension of their disfluencies. They go on to develop the characteristics of secondary stuttering: blocks, escape behaviors, and avoidance reactions. Why? And why do some children begin to stutter with blocks rather than repetitions? The answer, I think, can be found in the temperament of these children, interacting with the processes of learning.
A Perspective on Secondary Stuttering
In earlier sections I mentioned that several authors have suggested that the speeding up, tension, struggle, escape, and avoidance behaviors of secondary stuttering are a reaction to the simple repetitions of sounds and syllables that often characterize stuttering when it first begins. I now want to make the case that the child’s temperament and emotional responses interacting with learning can explain the behaviors of secondary stuttering.
This section focuses on the personality or temperament of children and uses both “sensitive” and “reactive” to refer to the same thing: a behavioral and emotional style characterized by being easily aroused by novel stimuli, as well as a tendency to withdraw when confronted by unfamiliar people or situations. Kagan (1994a, 1994b) often uses “inhibited” to describe the same traits. There is evidence, which was discussed in Chapters 2 and 3, that individuals who stutter tend to have more sensitive or reactive temperaments (Anderson, Pellowski, & Conture, 2001; Embrechts & Ebben, 1999; Fowlie & Cooper, 1978; Guitar, 2003; LaSalle, 1999; Oyler & Ramig, 1995; Wakaba, 1998). If so, such reactivity may explain why some children who stutter eventually respond to their disfluencies by tightening their muscles. Research on normal children who have sensitive temperaments indicates that they respond to novel, threatening, or unfamiliar events by increasing their physical tension, particularly in the larynx (Kagan, Reznick, & Snidman, 1987). If children who stutter have more sensitive temperaments than their peers, they may be more likely to increase their laryngeal tension in response to primary stuttering, which they experience as threatening because it seems out of their control. This may be the mechanism that causes many children’s disfluencies to change from easy, relaxed repetitions to tense repetitions and progressively more involved secondary behaviors.
The increases in physical tension just mentioned may be part of a larger defensive response that is triggered more easily in individuals with reactive temperaments. In describing his behavioral inhibition system, Gray (1987) proposes that when individuals experience frustration or fear, their innate response is freezing (i.e., widespread muscular contractions that produce tense and silent immobility), flight (i.e., speeded up activity to escape), or avoidance. Gray (1987) indicates that these unconditioned responses may occur rapidly without intervening autonomic arousal, arising from the central nervous system substrate underlying the increased muscle tension, increased tempo, and escape behaviors and avoidance behaviors that characterize secondary stuttering. It is notable also that Gray suggests that the way that individuals’ behavioral inhibition systems affect behavior is influenced by temperament. The more reactive individuals are, the more they will engage in freezing, flight, or avoidance responses when experiencing fear.
Further evidence of a neurological substrate underlying the characteristics of secondary behavior is provided by the research of Davidson (1984), Kinsbourne (1989), and Kinsbourne and Bemporad (1984), which was described in Chapters 2 and 3. They propose that the right hemisphere is specialized for emotions that accompany avoidance, withdrawal, and arrest of ongoing behavior, whereas the left hemisphere is specialized for emotions that are associated with approach, exploration, and release of ongoing behavior. Thus, it can be argued that individuals who stutter and are more reactive are more prone to behaviors regulated by right-hemisphere emotions. This argument is supported by the findings of Calkins and Fox (1994) and Davidson (1995) who reported that sensitive children are right hemisphere–dominant for emotion. Their research may also explain why secondary behaviors develop in the forms they do in many stutterers. Those beginning stutterers who develop secondary behaviors may be more sensitive individuals whose innate defensive mechanisms, the behavioral inhibition system of Gray (1987), are triggered more easily because of their righthemisphere dominance for emotions. It is possible that some of the excess activity seen in the right hemisphere of those who stutter, described in Chapters 2 and 3, may be related to activation of behavioral inhibition. These stutterers, especially those whose behaviors at onset are characterized by tension and struggle, may not be likely to spontaneously outgrow their stuttering.
The notion of greater emotional reactivity in children who stutter is supported by a study of 65 children who stuttered and 56 children who did not by Karrass and colleagues (2006). They found that compared to nonstuttering children, children who stutter have greater emotional reactivity, less emotional regulation, and poorer attention regulation. These authors suggest that this combination of traits contributes to the development of stuttering in a “reverberant” fashion. Reacting to what I would term “primary stuttering,” these children have a strong emotional response to their primary stutters and they are unable to regulate this emotion. This in turn makes them stutter more (and more severely), and they have even stronger emotional responses to the more severe stutters and on and on.
The importance of emotion in secondary (persistent) childhood stuttering is also supported by recent genetic evidence. As I indicated in Chapters 2 and 3, mutations in several genes, including GNPTG, have been linked to persistent stuttering, and it is also known that this same gene influences the development of the cerebellum and the hippocampus (Kang et al., 2010). The cerebellum is known for its role in motor control and in emotional regulation (e.g., Schmahmann & Caplan, 2006). The hippocampus is also a structure which influences emotion and is a key component of Gray’s behavioral inhibition system, described above. Thus, many lines of evidence converge on the link between secondary stuttering and emotion.
I’ve described many of the temperamental and emotional factors related to secondary stuttering as though they were completely within the child. But of course the child interacts with a complex environment, and this influences temperament and emotion, which in turn influence stuttering. An example that comes to mind is a girl and her mother in a study of parent-child interaction that we published many years ago (Guitar, Kopff-Schaefer, Donahue-Kilburg, & Bond, 1992). We computed correlations between several variables in the mother’s talking and the girl’s primary stuttering and secondary stuttering. What we found led us to surmise that different aspects of the mother’s conversation affected the two types of stutters differentially. The mother’s speech rate was highly correlated with the girl’s primary stutters but not with her secondary stutters. On the other hand, the mother’s nonaccepting comments were highly correlated with the girl’s secondary stuttering but not her primary stuttering. We speculated that when her mother made nonaccepting comments to her, the girl reacted with negative emotion, triggering secondary stuttering. We also speculated that when the mother talked fast, the girl tried to keep up, but because of an inefficient speech processing system, she stuttered with the easy repetitions of primary stuttering.
It has been evident for some time that learning plays a major role in persistent stuttering–stuttering that continues beyond the primary stage and usually grows more severe. In the last section of Chapter 5, I described the major types of learning involved in stuttering, but I did not deal with the question of why some children seem more vulnerable to this learning than others. One answer to this puzzle was given by Brutten and Shoemaker (1967) in their influential book, The Modification of Stuttering. They proposed that individual differences in conditionability and in autonomic reactivity are a constitutional predisposition in stuttering. My own understanding of the role of learning, described in the following paragraph, is in agreement with theirs and adds some detail from recent literature about emotional learning.
The perspective given earlier that some children who stutter are more temperamentally reactive and thus are more prone to physical tension, rapid escape behaviors, and avoidance suggests that these children will also be more emotionally conditionable. That is, the physiology that predisposes these children to be temperamentally reactive involves the limbic system, especially the amygdala, which is that part of the limbic system that stores emotional memories. This is what makes learning for them under conditions of strong emotion so indelible.
Emotional events are etched into the brain more strongly than neutral ones. Think about how well you can remember what you were doing when you found out about a very exciting or upsetting event. Many people can remember vividly what they were doing when they heard about the planes crashing into the twin towers of the World Trade Center in New York City on September 11, 2001. The strength of emotional memories is enhanced even further in people with more reactive limbic systems, which may account for why some people suffer posttraumatic stress syndrome and others do not. Children with reactive limbic systems are more likely to react to the multiple repetitions of primary stuttering with tension, escape, and avoidance and are also much more likely to store their memory strongly (e.g., LeDoux, 2002). Such reactions and memories can snowball. The child’s natural fear response to a repetition that feels out of control is to tense his muscles. This increased tension soon makes the stutter last longer, which increases his feeling that he is helpless and triggers a bigger fear response. Stuttering experiences become even more traumatic as the child increases physical tension because of his initial negative emotion, which causes repetitions to become tense prolongations or blocks. These new, tense forms of stuttering can be more unpleasant to the child and can provoke expressions of anxiety and alarm from his parents. The child’s stronger feelings of “stuckness” and his concern about his parents’ alarm are likely to generate more activity in the limbic system, thereby creating stronger memories that will trigger these more tense forms of stuttering more quickly whenever he experiences primary stuttering. The child’s reactive amygdala mediates the storage of unpleasant memories of stuttering, largely on an unconscious level. At the same time, another part of the limbic system, the hippocampus, stores information about the situations in which stuttering occurs (e.g., who the child was talking to, what word was being said, where it happened). These contextual cues cause stuttering to spread rapidly from isolated experiences to more and more repeated experiences in similar contexts and eventually to many other situations.
Children with reactive temperaments are not only more likely to learn to increase tension when they anticipate or experience stuttering but are also more likely to engage in other components of the behavioral inhibition system. These include increases in tempo, other aspects of escape behaviors, and a wide array of avoidances. Thus, these children quickly develop secondary symptoms, such as eye blinks and changing words, to avoid or escape stuttering.
Emotional conditioning occurs rapidly in these children, but unlearning is a much slower and more difficult process. There is evidence that emotional memories are stored permanently, and even when new behaviors replace them, the original emotions and learned behaviors may reappear under stress (Ayres, 1998). When clinicians work with individuals who have secondary stuttering, they need to keep in mind the strength and persistence of behaviors learned through fear conditioning. New behaviors will have to be learned as new responses to the stimuli that elicited the old responses. Because emotional conditioning may produce cortical as well as subcortical changes in the brain, cognitive therapy may be a useful adjunct to behavioral therapy in older children and adults.
Two Predispositions for Stuttering
This view of primary and secondary stuttering proposes that there are two constitutional predispositions: one for primary stuttering and one for secondary stuttering. As may be evident, the most common occurrence is for a child to have a predisposition for primary stuttering that is resolved through neural maturation or reorganization–this accounts for the 70 percent or so of children who spontaneously recover. It is also possible for a child’s primary stuttering to continue into adulthood and for secondary behaviors never to emerge; these adults may simply be considered highly disfluent, rather than stutterers. Some children, of course, do acquire secondary behaviors in response to their primary stuttering, and I would hypothesize that these children have the second predisposition, a reactive temperament, which leaves them prone to the tension, escape, and avoidance behaviors that characterize secondary stuttering. I believe that neither of these predispositions is “all or nothing;” that is, a child may have a little or a lot of either. For instance, an adolescent may have a substantial amount of repetitions in his speech but only occasionally show tension, escape, or avoidance behaviors. Or a child may start stuttering suddenly at age 3, with severe stutters marked by struggle, tension, escape, and avoidance. Perhaps that child has only a little predisposition for primary stuttering but a substantial predisposition for secondary stuttering. This continuum for stuttering agrees with most clinicians’ observations that we see a wide range of severity, from mild to very severe, with some stutterers having little avoidance and others having a great deal. I also notice that outside the clinic, there are many individuals whose “stuttering” is so mild that they don’t recognize it in themselves and other lay persons don’t notice it either.
The possibility of two predispositions for stuttering may also shed light on such phenomena as neurogenic stuttering, which are the disfluencies that sometimes appear in persons with neurological diseases or injuries. The changes in the brain that may occur as part of a neurological problem may give rise to a dyssynchrony in speech and language production processing that is similar to that of primary stuttering. On the other hand, the changes in temperament that sometimes occur with brain injury (e.g., Kinsbourne, 1989) may, in a few cases, give rise to disfluencies that are more characteristic of secondary stuttering.
There is support in genetic research for two (or more) predispositions in individuals who do not naturally recover from stuttering. After an analysis of many children who stutter for some period of time in their lives, Ambrose, Cox, and Yairi (1997) concluded that persistent and recovered stuttering are not two different forms of the disorder. Both persistent and recovered stutterers appear to have genetic factors related to the onset of stuttering. Those children who persisted in stuttering (i.e., continued to stutter for more than three years), however, have additional genetic factors, related to the persistency of the disorder. It may be that the additional genetic factors gives rise to a reactive temperament, making it more likely that these children will become frustrated, have other emotional reactions to their stuttering, and thus develop secondary/persistent stuttering. This connection between a child reacting to his stuttering and the persistence of the stuttering is reflected in Van Riper’s beliefs that “…most children who begin to stutter become fluent perhaps because of maturation or because they do not react to their … repetitions, or prolongations by struggle and avoidance … [while] those who struggle or avoid because of frustration or penalties will probably continue to stutter all the rest of their lives no matter what kind of therapy they receive” (Van Riper, 1990).
Indeed, as Ambrose, Cox, and Yairi (1997) suggest, there may be more than two predispositions in stuttering. The factors that cause a child to react with frustration and fear to primary stuttering may also cause another child to react the same way to lack of intelligibility or difficulties in word finding, for example. Communicative failures may lead to anticipatory struggle (Bloodstein & Ratner, 2008), as described earlier in this chapter. However, no matter how many predispositions a child may have, the chance of his actually developing primary or secondary stuttering is enhanced or diminished by both developmental and environmental factors.
Interactions with Developmental Factors
In this section, I describe three ways in which aspects of children’s development may interact with the two predispositions to trigger or exacerbate stuttering. In Chapters 4 and 5, I argued that children’s physical, cognitive, emotional, and linguistic development may provide the extra demands on resources that precipitate stuttering or worsen it. Now I want to relate these demands to the two predispositions described earlier. You will see elements of the capacities and demands theory of stuttering in this section.
The first interaction is with the demands of language development and a predisposition for primary stuttering. Consider a child who begins to acquire speech with dysfunctional or inefficient speech and language networks. The functional plasticity of the child’s brain may allow these pathways to reorganize or repair themselves so that the child processes spoken language more efficiently as he strives to communicate. However, the exponential growth of the child’s speech and language at this very time may compete for cerebral resources, straining or exceeding the child’s capacity to handle both the demands of reorganization and advancing language at the same time. To see what this may be like, imagine yourself as a student who has let part of the semester slip by without studying. After bombing the first two exams, you resolve to reorganize your study habits and catch up, but just then, your professors decide to pile on even more work than before. Like the child, you may or may not be able to accommodate the professors’ increasing demands at the same time you are spending energy to reorganize.
A second interaction will be the maturation of the brain with a predisposition for primary stuttering. Some individuals will have an earlier maturation of the brain or a natural flexibility to respond to anomalies in the wiring for spoken language. Girls, for example, are more likely to recover from early stuttering, probably because of their inherently greater organizational plasticity and their more widely distributed language centers (Shaywitz, Shaywitz, Pugh, Constable, & Skudlarski et al., 1995). Some males may also be genetically endowed with more flexibility than average for reorganizing their cerebral circuitry and thus may recover more readily than others.
The third type of interaction will occur when a child has normal neural circuitry for spoken language but has a constitutionally inhibited temperament. Typical developmental challenges for most children include some frustration at not being able to speak as fast or with the same complexity as adults and older children in the family. The child may not only be frustrated but embarrassed at his inability to produce more advanced speech and language. Social-emotional development takes the child through some stressful times. All of these typical experiences may produce increased tension and avoidance behaviors associated with speech. Based on my clinical experience, I suspect that some children fitting this description might be hesitant to speak and may be referred for a stuttering evaluation but would not manifest the typical signs of stuttering. Their hesitancies may consist of long pauses, phrase repetitions, or both when their right- hemisphere proclivity toward avoidance, withdrawal, and arrest of ongoing behavior manifests itself while they are speaking. Such hesitancies may diminish in time as myelinization of connections between and within the hemispheres progresses and the left hemisphere has increasingly modulating effects on right hemisphere–regulated emotions.
Interactions with Environmental Factors
Here, I would like to consider the influence of the environment on anomalous speech and language neural networks (predisposition for primary stuttering) and on constitutional predispositions for inhibited temperaments (predisposition for secondary stuttering).
Interactions of Anomalous Neural Networks with Environmental Factors
As a child’s developing central nervous system adapts to the inherited or acquired differences in her neural substrates for speech and language, the environment plays a role through various listeners’ responses to the child’s emerging speech and language skills. Obviously, a child’s family will have the most opportunities to provide acceptance and support. The accommodations they can provide, such as slower speech rates, fewer interruptions, and dedicated one-on-one listening time, may foster adaptations of the child’s inefficient, dyssynchronous neural networks. At least this environment will not stress the child’s speech and language production system and will probably enable the child to develop her own adapted rate of speech and language output. In contrast, an environment with many interruptions, rapid conversational give and take, demands for recitations, and little time for the child to talk may “overdrive” the child’s immature speech and language production system, produce an excess of disfluencies, and inhibit the successful adaptation of the child’s system to its original anomalous wiring.
Interactions of Temperament with Environmental Factors
The work of Calkins (1994), Kagan and Snidman (1991), and others suggests that families can have a strong influence on temperament. As Calkins and Fox (1994) expressed it, “the child’s interactions with a parent provide the context for learning skills and strategies for managing emotional reactivity.” In addition, the environmental factors that I have called “life events” can also influence the development of temperament. As noted earlier, Kagan (1994b) suggested that certain life events could cause a child who is not particularly reactive to become more reactive and inhibited.
Implications for Treatment
A supportive environment and specific therapy approaches may ameliorate the conditions that I have just been describing. By judicious control of speech and language processing demands, the environment may support a child’s adaptive neuroplasticity, enabling him to improve the efficiency of his speech and language neural networks. Equally important is the family’s fostering of the child’s adaptation of inefficient neural networks. With appropriate models of slower speaking rates and pausing, the child may develop a rate of speech and language processing that allows him to synchronize the various components of spoken language.
Families may also help a child develop a less inhibited temperament by encouraging positive, assertive behaviors. Therapy, too, can help a child respond to disfluencies with fewer inhibitory responses. Active, positive treatment sessions often lead to improvements in a child’s confidence. Training in fluency skills can provide a child with many satisfying speaking experiences, thereby reducing fear of talking. Development of a slower speaking style and the use of proprioception (conscious awareness of movement) may help a child make the best of an inefficient speech production system. Confronting feared words and situations, reducing tension in stutters, and improving eye contact during speech may help shift a child’s characteristic emotional valence from “avoidance” (right hemisphere) to “approach” (left hemisphere). In the chapters on treating stuttering, I expand on this theme and suggest a variety of other ways to help individuals overcome or compensate for factors that predispose them to stutter.
Some recent research indicates another way in which treatment may affect the two predispositions for persistent stuttering. Brain imaging studies of adults before and after treatment (De Nil, Kroll, Lafaille, & Houle, 2003; Neumann, Euler, Wolff von Gudenberg, Giraud, & Lanfermann et al., 2003; Neumann & Euler, 2010) suggest that successful treatment is associated with activation of left-hemisphere areas that had not been active before treatment and reductions in activations of right-hemisphere areas that were highly active. These findings parallel evidence from treatment of nonfluent aphasic patients whose recovery was associated with the same pattern (M. Naeser, personal communication, January 5, 2005). Because this relocation of circuitry serving spoken language places it in areas very near to those used by nonstutterers in the left hemisphere (Neumann & Euler, 2010), activity may be less vulnerable to disruption by right-hemisphere emotions, thus affecting the mechanisms for both primary and secondary stuttering.
Accounting for the Evidence
Let us now turn to the research findings and clinical observations for which these views of stuttering must account.
Stuttering Occurs in All Cultures
The fact that stuttering is universal should not be unexpected because it depends less on culture than on basic biological variations of the human brain. Many other disorders, such as dyslexia and specific language impairment, as well as such personality differences as sensitive temperament, are associated with atypical activity of the central nervous system and are also universal.
Stuttering is a Low-Incidence Disorder
The fact that the prevalence of stuttering is relatively low may be a consequence of chronic stuttering resulting from a combination of at least two biological predispositions, the co-occurrence of which does not happen frequently.
Stuttering Does Not Begin with the Onset of Speech
Why does stuttering usually begin only after fluency at the one- and two-word stage has been achieved? My view is that stuttering emerges first from disruptions caused by a child’s inefficient neural networks for speech and language processing. Perhaps the task of coordinating all the phonetic, phonological, syntactic, and semantic components of longer utterances is too much for inefficient neural circuitry under stress. Just like normal disfluency, the neural processing circuitry of children who stutter may be adequate to handle one-word utterances. But once children begin to reorganize their language functions from a lexical to a grammatical-rules basis and try out more complicated syntax, their inefficient neural organization breaks down. An added demand on their planning system is that the shift from one- to two-word utterances requires the use of a more complex prosody.
Remember that Kent (1984), Perkins, Kent, and Curlee 1991), and Wingate (1988) suggested that a major source of breakdown is in timing linguistic and paralinguistic components. This demand for complex prosody at the multiword stage at the same time that phonological, syntactical, and lexical demands are added is likely to be a time when an inefficient speech and language system cannot keep up with the demands for rapid and complex speech production. It is as if our imaginary bicycle factory described earlier in this chapter had been producing old style bikes with pedal brakes and no gears (the one-word stage) but is now being asked to produce bicycles with hand brakes and two sets of gears (multiword stage), even as customer demand is requiring faster work on the production line (speech rate increases with longer utterances).
Stuttering Sometimes Begins with Tense Blocks, But Often with Repetitions
There are children who begin to stutter with tense blocks that did not follow a period of repetitions and occasional prolongations. I have recently been working with a 2year-old girl who showed excessive squeezing and tension in her stutters after only a few hours of stuttering in a repetitive pattern. For most children who stutter, however, tension responses, as well as escape and avoidance reactions, are elicited by the frustration and fear provoked by early stuttering. Speech itself is not threatening, but when a long repetition occurs as a result of the dyssynchrony in the speech and language production system, the child feels that her speech mechanism is out of control. This triggers her tension, escape, and avoidance responses, and as this happens more frequently, learning takes place, and her tension, escape, and avoidance responses are soon triggered without “runaway” repetitions.
Not All Stutterers Have Relatives Who Stuttered
How do we account for both the genetic transmission of stuttering and that evidence of genetic transmission is lacking in some cases? Genetic transmission of stuttering in many cases may be through the two factors I just described: anomalous neural organization for speech and sensitive temperament. In some cases of childhood stuttering, genetic transmission may seem unlikely because no other family members seem to be affected. However, it may occur because persistent stuttering appears to require both predisposing factors. Some family members may inherit one factor and some the other, but unless both factors are inherited by the same individual, persistent stuttering does not develop. Another reason for the absence of stuttering in other family members may be that the predisposing factors were the result not of genetic inheritance, but of environmental factors affecting fetal development that created the neural substrate for stuttering. Moreover, such anomalous speech and language circuitry may create language, learning, or phonological problems in other family members. Remember that the unfolding of the genetic blueprint is extensively influenced by environmental factors and by chance. Thus, the anomalous circuitry in one child may result in stuttering, but in an uncle or grandmother, it may have resulted in an articulation disorder or learning problem.
Stuttering Appears as Repetitions, Prolongations, and Blocks
The immediate causes of the core behaviors of stuttering are not entirely clear in my view. All of them reflect an inability to move forward in speech, but the effortless sound and syllable repetitions of many stutterers at onset seem somewhat different from later tension-filled repetitions, prolongations, and blocks. The sound and syllable repetitions of early stuttering more closely resemble the disfluencies resulting from nervous system damage or “neurogenic stuttering” (Rosenbek, 1984). Thus, these early signs of childhood stuttering (less tense repetitions and prolongations) may arise from a breakdown in the inefficient function of neural circuits, perhaps from causes similar to those of neurogenic stuttering. Repetitions may occur simply because there is a lag in the readiness of the next part of a word or sentence, although the impulse or pressure to continue speaking is strong. Signs of stuttering with tension that emerge later than effortless repetitions in many stutterers likely stem from the frustration and fear elicited by a child’s difficulty in speaking. In those cases in which the earliest signs of stuttering are characterized by tension and blocking (Van Riper, 1982), an emotional response may be primary. As Van Riper (1982) suggested, these may be children whose onset is very sudden, resulting usually after an emotionally difficult period or traumatic emotional stress.
Stuttering is More Common in Boys than in Girls
I suspect that the reason more boys stutter than girls is that their genetic blueprints for neural organization of speech and language differ between boys and girls and may be more flexible in females (Shaywitz et al., 1995). Neuroplasticity of the human brain is greatest in the first few years of life, and this neuroplasticity probably diminishes after puberty. Neuroplasticity permits reorganization of neural pathways and, in many cases, recovery. Karlin (1947) advanced another explanation of why more girls recover early from stuttering. He postulated that delayed myelinization of nerve fibers in speech processing areas was a possible explanation of stuttering and cited research that myelinization of nerve fibers is more advanced in girls than in boys.
In Many Children, Stuttering Starts as Mild and Develops into a More Severe Form
The course of development of stuttering seems to us to be determined in part by the biological responses of the child to fear and frustration and to autonomic conditioning, to which a child prone to chronic stuttering may be particularly sensitive. Details on the development of stuttering are discussed in Chapter 7.
Many Conditions Reduce or Eliminate Stuttering
Conditions that temporarily ameliorate stuttering, such as singing or speaking in a rhythm, probably improve fluency by giving speech and language processes more time or an external organizing stimulus to aid speech production. They may also involve other parts of the brain and not those anomalous networks used inefficiently for typically spoken language. Other conditions, such as speaking when alone or when relaxed, often reduce stuttering but do not necessarily eliminate it. Such conditions may calm a person, thereby diminishing the reactivity of limbic circuits, whereas some conditions, such as speaking more slowly, both provide more time and diminish reactivity.
Stutterers Often Have Poorer Performance on Sensory and Motor Tasks
How about differences in performance between groups of stutterers and nonstutterers? As intimated earlier in this chapter, it seems to me that the wide range of performance on language tests, school achievement tests, and tests of sensory-motor ability by groups of stutterers may reflect the wide range of delays and deviations in the neural substrates for these abilities that led to their inefficient processing of speech and language. Among groups of individuals who stutter, there are likely to be some whose neural organization for sensory-motor processing is deviant enough to depress the group mean. In highly selected groups of stutterers, however, such as all males with no medical, neurological, or psychiatric diagnoses and who are right-handed (Ingham, Fox, Ingham, Zamarripa, & Martin et al., 1996), the chance of finding significant differences between this group and a group of nonstutterers is decreased.
In this regard, it is interesting that two independent studies have shown that children whose stuttering is their only disorder show no speech reaction time differences from nonstutterers, whereas children who stutter and have other language or articulation disorders show significantly poorer reaction time scores than children who only stutter and children with typical speech (Cullinan & Springer, 1980; Maske-Cash & Curlee, 1995). Perhaps the coexistence of poorer sensory-motor integration performance and speech and language disorders reflects additional anomalies in neural organization and function in this subgroup of stuttering children. In other words, I propose that children who stutter have at least some degree of inefficient organization of their neural circuitry for speech and language production; those children who stutter and have poorer sensory-motor skills or other speech and language disorders may simply have greater anomalies in their neural circuitry functions, which affect fluency, articulation, language, or other sensory-motor tasks.
Other characteristics of stuttering that I have said should be explained by any view of stuttering, such as the influence of developmental and environmental factors, are explicitly addressed in earlier parts of this chapter. Some characteristics, findings, and observations are explained more easily than others. However, those that are not accounted for in detail (such as the strong effect of rhythmic stimuli on stuttering) should not be ignored. They are a reality and are hard facts that should mold and shape any theoretical view until it is more fully explanatory.

• Several theoretical perspectives have been proposed to account for constitutionalfactors in stuttering. They include views of stuttering: (1) as an anomaly of how the brain is organized for speech and language; (2) as a disorder of timing of the sequential movements for speech; (3) as a result of deficits in the internal modeling process used to control speech production; (4) as a disorder of spoken language production; and (5) as a result of physiological tremor in speech musculature. The first four of these views focus on dysfunctions of cortical and subcortical mechanisms that control the planning and production of speech and language to produce the initial repetitions and prolongations of early stuttering; the fifth view targets neuromuscular malfunctions that may explain the tension and tremors of secondary stuttering.
• Theories concerning developmental and environmental factors include (1) the diagnosogenic theory, which implicates the listener’s response to the disfluencies of the child; (2) the anticipatory struggle theory, which suggests that a child may develop stuttering as a result of negative anticipation of speaking after he has had frustrating or embarrassing experiences in communicating; and (3) the capacities and demands theory, which postulates that stuttering arises when the child’s capacities for rapid, fluent utterances are unequal to the demands within the child himself or within the environment.
• In this chapter, I elaborate a two-stage etiological model of stuttering that I firstproposed in a chapter on children’s stuttering and emotions (Guitar, 1997) and that owes much to Bluemel (1957) and Brutten and Shoemaker (1967). The first stage is primary stuttering, which involves repetitions and prolongations that are frequently the first signs of stuttering. These signs are thought to be the result of a constitutional factor: a dyssynchrony at some level of the speech and language production process. The second stage is secondary stuttering, which involves the tension, struggle, escape, and avoidance behaviors that are often present in persistent stuttering. These behaviors are proposed to be the result of a separate constitutional factor–a reactive temperament that triggers a defense response from the behavioral inhibition system and that makes the individual more emotionally conditionable than the average speaker.

1. What are the differences between the Geschwind and Galaburda (1985) theory ofstuttering and the Webster (1993a) view?
2. Compare Kent’s (1984) view of stuttering as a disorder of timing with the Geschwind and Galaburda (1985) theory.
3. Both Neilson and Neilson’s view of stuttering and one of Max and colleagues’(2004) hypotheses about stuttering suggest that repetitions occur because of a problem with the internal models used for speech production. What is the difference between the cause of repetitions in each view?
4. The study by Kelly , Smith, and Goffman (1995) reviewed in this chapter suggested that tremors don’t appear in younger children who stutter but do appear in older children. Why would this be?
5. Table 6.2 lists experiences that may generate stuttering in some children because the experiences have led children to believe speaking is difficult. Add as many other hypothetical experiences as you can to this list.
6. A capacities and demands view of stuttering in children would lead to a therapystrategy of enhancing a child’s capacities. What are some examples of what capacities in a child you could strengthen to reduce stuttering? Describe how you would do this.
7. What is the relationship between a sensitive temperament and a high level ofconditionability?
8. I have suggested there may be two predispositions for persistent stuttering–onefor primary stuttering and one for secondary stuttering. How, according to this view, would primary stuttering lead to secondary stuttering?
9. There is strong evidence that girls are more likely than boys to recover fromstuttering and are therefore less likely to become persistent stutterers. Is this because girls are more likely to recover quickly from primary stuttering or because their primary stuttering is less likely to trigger secondary stuttering?
10. In this chapter, I have suggested that there are two stages of stuttering–primaryand secondary. What are the signs and symptoms of each, and what is the suggested etiology of each?
1. The view of stuttering as a problem of the “internal modeling” process inspeech production is a complex idea. Read the article by Max and colleagues (2004) and make a class presentation about their full theoretical model, explaining it in as clear and simple a way as possible.
2. Read the article entitled “Resources–A Theoretical Stone Soup” (Navon, 1984) and use the arguments in it to evaluate the capacities and demands theory in this chapter.
3. Wendell Johnson’s “diagnosogenic” view of stuttering led to a master’s thesisthat tried to create stuttering in orphans in 1939. In 2003, this thesis was the topic of a controversy that centered on the ethics of trying to induce stuttering
in children. Using the Internet, research this controversy, using “Monster Study” as a keyword. Make a presentation or write a paper on the ethics of this research, given the fact that it was conducted more than 50 years ago when the ethical climate was markedly different than it is now.
4. Pick a theory of stuttering–either one described in the first two sections ofthis chapter or one you have found elsewhere–and evaluate how it can account for the basic facts about stuttering enumerated in Chapter 1.
5. Go to Guitar and McCauley (2010) and read two chapters concerning specificinterventions related to a specific age group of people who stutter. Identify which theories addressed here are cited and how they appear to impact the developers of each intervention approach.
Brutten, E. J., & Shoemaker, D. J. (1967). The modification of stuttering.
Englewood Cliffs, NJ: Prentice-Hall, Inc.
This is a classic book in the field of stuttering. It describes a theory of stuttering that ascribes the initial symptoms of childhood stuttering to the effect of anxiety on fluency and ascribes the later symptoms to learning. The authors go on to suggest therapeutic approaches that derive from their model.
Guitar, B., & McCauley, R. (2010). How to use this book. In B. Guitar, & R. McCauley (Eds), Treatment of Stuttering. Baltimore: Lippincott Williams & Wilkins.
This chapter describes in user-friendly language what a theory is and how it can help researchers and clinicians.
Gray, J. A. (1987). The psychology of fear and stress (2nd ed.). Cambridge: Cambridge University Press.
Gray’s experimental work and his theoretical model of a behavioral inhibition system are clearly described here. Some of the book (those parts dealing with the effects of pharmacological agents on the brain) is for specialized readers. Much of it, however, is a readable exposition on the biological basis of learning, stress, and fear. Because this is the second edition of a popular book, I hope a new edition will be available soon.
Kagan, J., Reznick, J. S., & Snidman, N. (1987). The physiology and psychology of behavioral inhibition in children. Child Development, 58, 1459– 1473.
This article discusses the findings of Kagan and his colleagues that behaviorally inhibited children show high levels of laryngeal tension. Neurophysiological mechanisms are also discussed, as well as possible genetic and environmental contributions. This book is recommended for those interested in the hypothesis that behavioral inhibition may be a component in some stuttering.
LeDoux, J. (1996). The emotional brain: The mysterious underpinnings of emotional life. New York: Simon & Schuster.
LeDoux, a highly respected brain researcher, brings together a great deal of evidence about how the brain processes experiences that we consider emotional.
His explanations of emotional learning are very clear and relevant to stuttering.
Packman, A., & Attanasio, J. (2004). Theoretical issues in stuttering. New York: Psychology Press.
The authors review current and past theories of stuttering and evaluate them in terms of testability, explanatory power, parsimony, and heuristic power. This book effectively teaches the reader what a theory should be expected to do.
1 This notion of the left hemisphere being specialized for more rapidly changing signals has recently been challenged by Boemio, A., Fromm, S., Braun, A. & Poeppel, D. (2005).
2 See Guitar, Guitar, Neilson, O’Dwyer, and Andrews, 1988 for early evidence in linewith Smith and Kelly’s view. This study found that electromyographic signals reflecting muscle contractions closely reflected whether moments of stuttering were clearly audible vs. subtly present versus not evident at all.
Normal Disfluency and the Development of

Exceptions and Variations
Normal Disfluency
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Underlying Processes
Younger Preschool Children: Borderline Stuttering
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Underlying Processes
Older Preschool Children: Beginning Stuttering
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Underlying Processes
Increases in Muscle Tension and Tempo
Effects of Learning on Stuttering
School-Age Children: Intermediate Stuttering
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Underlying Processes
Older Teens and Adults: Advanced Stuttering
Core Behaviors
Secondary Behaviors
Feelings and Attitudes
Underlying Processes
After studying this chapter, readers should be able to:
• Describe and explain the ways in which the age and developmental levels haveexceptions and variations
• Describe and explain the (a) core behaviors, (b) secondary behaviors, (c) feelings and attitudes, and (d) underlying processes for the following age and developmental levels: normal disfluency
stuttering in younger preschool children: borderline stuttering stuttering in older preschool children: beginning stuttering stuttering in school-age children: intermediate stuttering stuttering in older teens and adults: advanced stuttering
Underlying processes: These are speculations about the process that may cause disfluencies or stuttering at each developmental level. For stuttering, these processes help us understand why stuttering often changes from borderline to beginning to intermediate to severe levels
Age/developmental levels: These levels reflect both the age of the individual (e.g., younger preschooler, older preschooler, etc.) and the severity of the stuttering (e.g., borderline, beginning, etc.)
Borderline stuttering: This is the earliest or lowest level of stuttering, usually seen in children ages 2 to 3.5. This type of stuttering is characterized by more frequent part-word and single-syllable whole-word repetitions than typically developing children have, but without awareness or concern on the part of the child
“Within-word” disfluencies: Disfluencies that occur within a word boundary such as repetitions of parts of words, prolongations, or blocks. Stuttered

speech is said to contain a higher proportion of within-word disfluencies (as opposed to disfluencies that happen between words or across words, such as hesitations, fillers, and repetitions of whole words). Note that disfluencies of typically developing children may also include within-word disfluencies
“Stuttering-like” disfluencies: Short-segment repetitions (i.e., part-word and monosyllabic word repetitions), as well as sound prolongations, and blocks. These are disfluencies that are more typically experienced by listeners as associated with stuttering
Dysrhythmic phonation: A sound prolongation, broken word, or other instance of ongoing phonation being stopped, extended, or distorted
Beginning stuttering: This level of stuttering is usually seen in children between ages 3.5 and 6, although it may occur before and after those ages. It is characterized by more tension and hurry in disfluencies than that seen in borderline stuttering. Stuttering usually consists of repetitions and prolongations, but some children will also exhibit blocks. Escape behaviors appear in this level of stuttering
Intermediate stuttering: Typical of children in their school-age years, this level of stuttering will abound in repetitions and prolongations, but blocks will also be frequent. In addition to escape behaviors, avoidances will appear at this level because there is fear of being “stuck” in a stutter and fear of listener reactions
Starters: Words or sounds used by someone who stutters to get started speaking when blocked or when anticipating a block. For example, a person who stutters might say “My name is, uh, Barry”
Substitutions: The substitution of an “easier” word for a “harder” word on which a stutterer expects to stutter. For example, a stutterer who often stuttered on words beginning with “p” and who had a dog named “Pluto” might always substitute “my dog” for the dog’s name when talking about him
Circumlocutions: Rather than stutter on a word, a person who stutters might use a different way of saying something, such as “My father was in the N…n…he served aboard ships in the armed forces”
Postponements: This is like a starter, but usually it just involves waiting a few beats before saying a feared word as in “Back then I voted for Ronald……… Reagan”
Antiexpectancy devices: An unusual way of speaking or acting that seems to reduce stuttering, like laughing and pretending that most things said were a joke. Another example is speaking with an accent that the speaker pretends to have
Avoidance conditioning: A type of learning that occurs when a person avoids something he thinks will be unpleasant. The avoidance is rewarded by the fact that the unpleasantness doesn’t happen. Avoidance conditioning is important
in thinking about stuttering development and interventions because it can be difficult to combat
Advanced stuttering: This level is characteristic of older teens and adults who have been stuttering since childhood. Their stuttering pattern is quite ingrained, especially behaviors associated with avoidance and ways of coping with blocks
This chapter describes the development of stuttering and what it looks like at various ages. It is designed to help you understand why once stuttering has emerged, it often (but not always) progresses from a few relaxed repetitions in preschool children to frequent stuttering accompanied by tension, avoidance, and many negative feelings and beliefs in older children or adults. This chapter will also help you understand how to match treatment procedures to the underlying dynamics of stuttering, as well as to the age of the client. To accomplish these goals, I have organized the content into five levels that reflect not only age groupings but also stages of development of stuttering and important characteristics of each stage of stuttering to guide your selection of a therapy approach (Fig. 7.1)

Figure 7.1 Overview of Chapter 7.
The five age groupings/developmental levels are given in Table 7.1, along with the four subcategories of stuttering characteristics. The first three subcategories–core behaviors, secondary behaviors, and feelings and attitudes–were described in a general way in Chapter 1. The fourth, underlying processes , is introduced to explain why symptoms may change from level to level. These explanations are hypotheses based on evidence from studies of animal and human behavior about how stuttering behaviors become more severe and complex. This subcategory should help you understand the nature of the symptoms as well as the rationales for the treatments presented in the second section of the book.
Table 7.1 Developmental/Treatment Levels of Stuttering

Specific age groupings (younger preschool, older preschool, school-age, and teens and adults) are used because age is often critical in selecting the appropriate treatment. Let me give two examples. No matter how severely a preschool child stutters, treatment should always involve his parents, and in my experience, it should focus primarily on increasing fluency rather than modifying stutters. Conversely, a schoolage child–whether stuttering is mild or severe–needs an approach that involves teachers and classmates as well as parents. Also, it helps children of this age to discuss their stuttering and their feelings about it. In general, the cognitive-emotional level typically associated with different ages needs to be considered when choosing a therapy strategy.
Exceptions and Variations
The age/developmental levels presented in this chapter do not characterize everyone who stutters. For example, some older preschool children may be stuttering so mildly and be so relatively unaware of it that they might best be treated by an approach described for younger preschool children. However, most individuals who stutter will fit reasonably well into their age grouping. Moreover, even though all behaviors of a person may not reflect a single age/developmental level, deciding on a treatment need not be a problem; when some aspects of a person’s stuttering seem more advanced than others, strategies can be borrowed from other levels to treat them.
Another qualification of the hierarchy presented here concerns the implication that all individuals who stutter pass through each stage in sequence. This is generally true, but there are exceptions. A child may show only normal disfluencies one day and beginning or intermediate stuttering on another day. She may stop stuttering without apparent reason a week later, or she may continue stuttering unless treated. One 3year-old boy I knew changed overnight from borderline to severe beginning stuttering after a change in his allergy medication. As soon as he resumed taking the original prescription, he became a borderline stutterer again and then recovered completely without treatment. There are many unsolved mysteries in stuttering.
Two clinical researchers who wrote extensively about the development of stuttering, Van Riper (1982) and Bloodstein (1960a, 1960b, 1961a), agreed that a simple sequence of stages could never capture every stutterer’s pattern. Bloodstein (1960b) proposed a series of four stages of stuttering development, which he described as “typical, but not universal” (Bloodstein, 1995, p. 53). He also cautioned that although stuttering near onset is often characterized by repetitions without awareness or by a lack of concern, some children at this stage show considerable effort and strain in their stuttering as well as crying from frustration at their inability to produce speech easily (Bloodstein, 1960a).
Van Riper (1982) also noted the presence of forcing and struggle in some children at the onset of stuttering, and like Bloodstein, he was struck by the fact that most children, especially in their early years, oscillate between remissions and recurrences of their stuttering, between mild stuttering and normal disfluency, or between more advanced and less advanced stages of development.
In addition to such swings in the progression of stuttering development, there may also be different paths of development, which different stutterers may follow. After searching his clinical files on many individuals whom he had followed for several years, Van Riper (1982) found that his data suggested there are subgroups of stutterers who are characterized by different onsets and different trajectories of development. He proposed that there are four distinctive “tracks” that an individual may follow. The most common consists of children whose stuttering begins as repetitions between 2 and 4 years of age, progresses to include prolongations, then gradually develops into blocks with more and more tension as well as fears and avoidances. The next most common track comprises children whose onset is a little later and is sometimes accompanied by delayed speech development, articulation problems, or very rapid speech. An interesting aspect of this track is that these children seem to have had difficulty hearing their own speech, perhaps as a result of auditory processing problems. This is particularly interesting in light of recent findings from brain imaging studies of adults who stutter (e.g., Foundas et al., 2001), indicating that some have anatomical anomalies that might produce difficulty with higher-level auditory processing. A third less common track includes children who have sudden onsets of stuttering with a great deal of tension that results in tight, laryngeal blocks. Finally, a fourth track consists of individuals whose disfluency appears to have psychogenic components. (An expanded discussion of psychogenic disfluency is presented in Chapter 15.) This track is characterized by late onset, a stereotyped pattern of stuttering that changes very little with age, and few avoidances. Van Riper’s four tracks serve as a warning to us that there is much diversity in the evolution of stuttering.
Keeping these variations, exceptions, and limitations in mind, I now begin a detailed description of the levels of stuttering development and treatment, starting with a group of behaviors that is really not stuttering at all, but a part of normal speech.
Children vary a great deal in how disfluent they are as they learn to communicate. Some pass their milestones of speech and language development with relatively few disfluencies. Others stumble along, repeating, interjecting, and revising as they try to master new forms of speech and language on their way to adult competence. Most are somewhere between the extremes of exceptional fluency and excessive disfluency, such as the 2-year-old shown in Figure 7.2.

Figure 7.2 Child who may be normally disfluent.
Children also swing back and forth in the degree of their disfluency. Some days they are more fluent and other days less fluent. Such swings in disfluency may be associated with language development, motor learning, or other developmental or environmental influences mentioned in the preceding chapters. In the following sections, I discuss factors that may influence disfluency, specific behaviors that I categorize as normal disfluency, and the reactions that some children may have to their disfluency. I also highlight aspects of normal disfluency that distinguish it from early stuttering, because one of my aims in this chapter is to prepare you to make this differentiation.
Core Behaviors
Normal disfluencies have been cataloged by several authors, and there is general agreement among them about what constitutes disfluency (Bloodstein, 1987; Colburn & Mysak, 1982a, 1982b; Williams, Silverman, & Kools, 1968; Yairi, 1982, 1983, 1997a; Yairi & Ambrose, 2005). Table 7.2 lists eight commonly used categories of disfluency.
Table 7.2 Categories of Normal Disfluencies

Some of the major distinguishing features of normal disfluency–features that differentiate it from stuttering–are the amount of disfluency, the number of units of repetitions and interjections, and the type of disfluency, especially in relation to the age of the child.
Let’s begin with the amount of disfluency. This is often measured as the number of disfluencies per 100 words or syllables, rather than “percent disfluencies.” Percent disfluencies implies that the disfluencies are associated with the production of particular words. For example, if you said that a child had 10 percent disfluent words, it would be assumed that 10 percent of the words spoken were spoken disfluently. However, many disfluencies, such as revisions, interjections, or phrase repetitions, are associated with several words or occur between words. For example, a child may say “Mommy, can you … can you … um … can you buy me that?” It’s inaccurate to say that some of these words were spoken disfluently, because the disfluencies were the repetition of the phrase “can you” and the interjection of “um.” Were the disfluencies on the words spoken or did they occur because the child was having trouble formulating the remainder of the sentence? In this case, we say that the child spoke six words ( “Mommy can you buy me that?”) and had two disfluencies (a phrase repetition and an interjection). Hence, we calculate the number of disfluencies that occur when the child speaks 100 words. More details on counting disfluencies are given in Chapter 8.
Although many researchers have measured disfluencies per number of words spoken, a good argument can be made for measuring disfluencies per number of syllables spoken. Andrews and Ingham (1971) first recommended the practice of assessing frequency of stuttering in relation to syllables spoken because some multisyllable words may have more than one disfluency, like “S-S-S-Sept-t-t-tember” or “di-dinosa-sa-saur.” These examples would be one disfluency each if disfluent words were counted, but two if disfluent syllables were counted. In line with this, Yairi (1997a) noted that as children get older, they are more likely to use multisyllable words. To keep the count equitable between younger and older children, Yairi has assessed disfluencies in children as the number per 100 syllables attempted (Hubbard & Yairi, 1988; Yairi & Ambrose, 1996; Yairi & Lewis, 1984).
When the frequency of all of a child’s disfluencies is measured, we need to know how many of these disfluencies are normal. Some of the earliest research on disfluency was conducted by Wendell Johnson at the University of Iowa. He assembled a team of researchers in the 1950s to examine the evidence for his “diagnosogenic” theory of stuttering. As indicated in Chapter 4, Johnson hypothesized that at the time a child is first “diagnosed” a stutterer by his parents, the child’s disfluencies do not differ from those of nonstuttering children. One of the research team’s projects was to record children identified by their parents as stutterers and compare the disfluency in their speech with that of nonstuttering children (Johnson & associates, 1959). One part of this study compared 68 male children who stuttered with 68 male children who didn’t. The results showed that although there was some overlap, the stuttering children had more than twice the amount of disfluency (on average, 18 disfluencies per 100 words) than did the nonstuttering children (only seven disfluencies per 100 words). Johnson interpreted the findings as showing that the two groups were essentially the same because there was so much overlap in both amount and type of disfluency. Other researchers (e.g., McDearmon, 1968) have reinterpreted these data as indicating there are two different groups, as I discussed in the last chapter.
Other researchers who have examined the disfluencies in nonstuttering children put the amount of their disfluencies at about the same level as Johnson and his colleagues reported (DeJoy & Gregory, 1985; Hubbard & Yairi, 1988; Wexler & Mysak, 1982; Yairi, 1981; Yairi & Ambrose, 1996; Yairi & Lewis, 1984; Zebrowski, 1991). Bringing all these studies together, we can estimate that normally speaking preschool children have on average about seven disfluencies for every 100 words spoken. If measured in terms of syllables, it would be closer to six disfluencies per 100 syllables. This figure may be a little high if children are examined throughout their preschool years (Yairi, 1997a); however, many children at age 2 or 3 go through a period of increased disfluency, which will reach this level.
The range in frequency of normal disfluency is important to note also, especially if the frequency of disfluency is used to make clinical decisions. Johnson and associates (1959) and Yairi (1981) found that, although many nonstuttering children have only one or two disfluencies per 100 words, at least one child in their samples had slightly more than 25 disfluencies per 100 words. Thus, the frequency of disfluencies is not a definitive clinical measure by itself.
Another distinguishing characteristic of normal disfluency is the number of units that occur in each repetition or interjection. Yairi’s (1981) data suggest that normal repetitions typically consist of only one extra unit. For example, a child might say “That my-my ball.” Interjections are likely to be just a single unit, such as “I want some … uh … juice.” Instances of multiple repetitions were occasionally observed in these children, but they were the exception. The rule is one and sometimes two units per repetition or interjection, which agrees with the findings of Johnson and associates (1959) that average nonstuttering children have one- or two-unit repetitions.
Another major characteristic of normal disfluency is the type of disfluency that is most common. Johnson and associates (1959) found that interjections, revisions, and whole-word repetitions were the most common disfluency types among the 68 nonstuttering males, who ranged in age from 2.5 to 8 years of age. Yairi’s (1981) study of 33 typically developing 2-year-old children found that there were two clusters of common disfluency types. One cluster involved repetitions of speech segments of one syllable or less (one-syllable words or parts of words were repeated). The second cluster consisted of interjections and revisions.
The most common disfluency type seems to change as a child grows older. In a follow-up to his earlier study, Yairi (1982) found that as children matured between 2 and 3.5 years, they gradually increased their frequency of revisions and phrase repetitions but decreased their frequency of part-word repetitions and interjections. He suggested that these data indicate that as nonstuttering children mature, part-word repetitions decline, even if other disfluency types increase. Thus, an increase in partword repetitions as a child is observed longitudinally may be a sign that warrants concern.
Although the research is far from complete, we can characterize normal disfluency types as follows:
• Revisions are common in normal children and may continue to account for a majorportion of their disfluencies as they grow older.
• Interjections are also common, but usually decline after 3 years of age.
• Repetitions may also be a frequent type of disfluency around 2 to 3 years of age,especially single-syllable word repetitions having fewer than two extra units. Repetitions are also more likely to involve longer segments (e.g., phrases) as a child grows older.
Table 7.3 summarizes the major characteristics of normal disfluency.
Table 7.3 Characteristics of Normal Disfluency in the Average
Nonstuttering Child

Secondary Behaviors
A normally disfluent child generally has no secondary behaviors. He has not developed any reactions to his disfluencies, such as escape or avoidance behaviors. Although research suggests that some normal children occasionally display “tense pauses,” such tension does not appear to be a reaction to their disfluencies. If a child shows what appears to be normal disfluencies, such as single-word repetitions, but consistently displays pauses or interjections of “uh” immediately before or during disfluencies, he should be carefully evaluated for possible stuttering.
Feelings and Attitudes
A normally disfluent child rarely notices his disfluencies, even though they may be apparent to others. Just as a child may stumble when walking but regains his balance and keeps walking without complaint, a typically developing child who repeats, interjects, or revises usually continues talking after a disfluency without evidence of frustration or embarrassment.
Underlying Processes
First, let’s review the behaviors for which we are trying to account. Normal disfluency occurs throughout childhood and adulthood. It may begin earlier than 18 months of age and peak between ages 2 and 3.5 years. It slowly diminishes, thereafter, but also changes in form. Some types of disfluency, such as repetitions, decrease after 3.5 years, but other types, such as revisions, may increase. Episodic increases and decreases in disfluency are also common throughout childhood. What causes these changes? Why are there ups and downs and changes in form? Like most natural phenomena, multiple forces probably have an impact on fluency at any given moment, but some may predominate at certain times. In Chapters 4 and 5, I talked about developmental and environmental influences on stuttering and normal disfluency, and I will review these influences as I discuss studies of normally disfluent children.
Certainly, the development of language is likely to be one major influence on fluency. As my earlier review showed, children tend to be most disfluent at the beginning of syntactic units (Bernstein Ratner, 1981; Bloodstein, 1974, 1995; Silverman, 1974) and when the length or complexity of their utterances increases (DeJoy & Gregory, 1973; Gordon, Luper, & Peterson, 1986; Pearl & Bernthal, 1980). These findings suggest that disfluency is greatest when a child is busy planning long or complex language structures yet must, at the same time as he is planning, begin to produce them, a process that places a heavy load on cerebral resources. It seems likely that producing newly learned language structures would be hardest of all and could result in more frequent disfluencies on a child’s most recently acquired forms. However, evidence gathered from four children between 2 and 4 years of age suggests that normal disfluency may be greatest on structures that have been learned but perhaps not fully automatized, thereby requiring more cerebral resources than are allocated to their production (Colburn & Mysak, 1982a, 1982b).
Pragmatics may influence disfluency, too. Studies by Davis (1940) and by Meyers and Freeman (1985a, 1985b) indicate that children’s disfluency increases under certain pragmatic conditions, such as when interrupting, when directing another’s activity, or when responding to requests/demands to change their own activity. Mastering such pragmatic skills, especially those involving more complex social interactions, creates yet another challenge for a developing child. The pressures of language acquisition, interacting with other factors, can be seen as competing for cerebral resources, which leaves fewer remaining resources available for fluent speech production.
In addition to language acquisition, another likely influence on disfluency is speech motor control. As they mature between 2 and 5 years of age, most children learn to produce almost all the segmental and super-segmental targets of their native language, as well as to increase their speech rates as they produce longer and longer utterances. These maturational changes must keep the average child fairly busy, although the demanding nature of these changes may not be obvious. The child is automatically scanning his parents’ and older siblings’ speech, acquiring information about talking. He is also continuously modifying his own productions to make them more and more like the speech he hears. This period–from 2 to 5 years–also encompasses an intensive refinement of nonspeech motor skills. It is at this age that children are learning to skip, run, jump, and take part in numerous games requiring skill and speed. Thus, children are mastering a myriad of other motor tasks at the same time they are acquiring the ability to speak in rapid, complex, fluent sequences.
In general, the view of stuttering described in Chapter 6, suggesting that a breakdown in speech fluency may occur when some of the neural pathways critical for sensorimotor control of speech production are immature or inefficient, may fit a nonstuttering child as well. Remember the dyssynchrony hypotheses we discussed in Chapter 6? Even normal disfluencies may be the product of a child learning to integrate all the subcomponents of spoken language at increasingly faster rates with increasingly greater options for vocabulary, syntax, and prosody.
Besides the continuing demands of normal development, there are also episodic stresses in a child’s environment that may temporarily increase normal disfluency. An experiment by Hill (1954) demonstrated that conditioned fear could elicit disfluency in normal adults’ speech. It is easy to imagine, therefore, that there are many psychological stresses in a child’s life that would also increase disfluency. Clinically, I have observed many situations that seem to increase normal disfluency. Among them are the stress of a move from one home to another, parents’ separation or divorce, the birth of a sibling, and other events that may decrease a child’s sense of security.
We have also seen increases in normal disfluency during periods of excitement, such as holidays, vacations, and visits by relatives. Disfluency increases especially when excitement combines with competition to be heard, such as during dinner table conversations when everyone is talking at once or after school when several children are competing to tell Mom what happened during the day. As we speculated in Chapter 6, emotions may have an especially strong influence on fluency in young children. This happens after interactions between right and left hemispheres develop during the child’s first two years (Fox & Davidson, 1984), and overflow activity from emotional arousal in the right hemisphere may disrupt vulnerable, immature language production networks in the left.

A video clip with an example of a child with normal disfluency is available on thePoint.
1. Between ages 2 and 5, many children pass through periods of increased disfluency.Repetitions, interjections, revisions, prolongations, and pauses are commonly heard during this period.
2. When the average child is between 2 and 3.5, disfluencies reach seven per 100 words spoken and may occur even more frequently in some normally disfluent children.
3. Repetitions are probably the most common type of normal disfluency in youngerchildren, whereas revisions are more common normal disfluencies in older children.
4. Despite the fact that children’s disfluencies may occasionally attract some adultattention, normally disfluent children seem generally unaware of the disfluencies in their own speech and don’t react to them or engage in secondary behaviors to escape or avoid them as a consequence.
5. Some factors thought to contribute to increases in normal disfluencies include thedemands of language acquisition, inefficent speech-motor control skills, interpersonal stress associated with growing up in a typical family, and threats to security from such events as relocation, family breakup, or hospitalization. Disfluencies may also increase under the ordinary daily pressures of competition and excitement while speaking.
Stuttering in preschool children between the ages of 2 and 3.5 resembles normal disfluency, but differs in several important ways. The most obvious–the thing that gets parents’ attention–is that these children have more disfluencies. We will discuss other key differences in the following sections. Sometimes diagnosis is difficult, because a child may drift back and forth between normal disfluency and borderline stuttering over a period of weeks or months. Some children with borderline stuttering gradually lose their stuttering and grow up without a trace of stuttering. Others develop more stuttering symptoms and progress through levels of beginning, intermediate, and advanced stuttering. Still others may continue to show borderline stuttering throughout their lives but may never seek treatment because their disfluency is so mild. A speech sample of a younger preschool child with borderline stuttering is depicted in Figure 7.3.

Figure 7.3 Child who may be a borderline stutterer.
In describing the behaviors of borderline stuttering, I will begin to define my view of how stuttering differs from normal disfluency. The distinction between stuttering and normal disfluency has been of great interest to theorists for many years. Some theorists (e.g., Johnson, 1955; Johnson et al., 1942) suggested, as was noted previously, that a stuttering child developed symptoms only after his parents mislabeled his normal disfluencies as stuttering. That is, a child’s first “stuttering” symptoms were actually just normal speech disfluencies.
An opposing view maintains that there are objective differences between the speech of a normal child and the speech of a child who is stuttering, even before a parent or someone else labels the behaviors as stuttering. Although I hold this latter view, I also agree that there is much overlap between the disfluencies of stuttering children and the disfluencies of normally disfluent children. Moreover, as previously stated, these children often go back and forth between stuttering and normal disfluency over a period of months. For this reason, we use the term “borderline” to indicate that these children are neither entirely normally disfluent nor definitely stuttering.
Core Behaviors
There is no single core behavior that distinguishes borderline stuttering from normal disfluency. However, many researchers and clinicians have suggested three elements that are useful for making this distinction. The frequency of disfluencies is one important aspect to consider. As we indicated in our description of normal disfluencies, nonstuttering children between 2 and 5 years may go through periods of increased disfluency. Even so, their level of disfluency averages about seven per 100 words. Typically, if children have many more disfluencies per 100 words (e.g., more than 10), we consider them borderline.
Another feature that can help identify borderline stuttering rather than normal disfluency is the proportion of certain types of disfluencies. The study we cited earlier by Johnson and colleagues (1959) suggested that compared to nonstuttering children, those who stutter had significantly more sound and syllable repetitions, word repetitions, phrase repetitions, broken words (i.e., phonation or airflow is abnormally stopped within a word), and prolonged sounds. There were no significant differences between the groups in their number of interjections, revisions, or incomplete phrases.
More information on types of disfluencies was provided by Young (1984), who reviewed a large number of studies that had assessed which types of disfluencies were identified as stuttering and which were not. His summary impression was that repetitions of parts of words, and to a lesser extent prolongations, are the disfluency types that are likely to be classified as stuttering. Bloodstein and Ratner (2008) and Conture (1982, 1990, 2001) generally concurred with other writers, suggesting that “within-word” disfluencies (i.e., part-word repetitions and audible as well as inaudible prolongations including blocks) are the types of disfluencies most frequently heard in stuttering children.
Yairi and colleagues (e.g., Yairi, 1997a, 1997b; Yairi & Ambrose, 1996) proposed that children who stutter can be distinguished from normally disfluent children using a grouping of “stuttering-like” disfluencies. Included in this grouping were shortsegment repetitions (part-word and monosyllabic word repetitions); tense pauses (stoppage of speech with evident muscular tightening both within and between words); and a category introduced by Williams, Silverman, and Kools (1968) called “dysrhythmic phonation” (any distortion, prolongation, or break in phonation within a word). Yairi (1997a, 1997b) notes that when many previous studies of stuttering and nonstuttering children are reanalyzed using this grouping, the proportion of stutteringlike disfluencies in nonstuttering children is always less than 50 percent of the total number of disfluencies. Thus, if a child has more than 50 percent stuttering-like disfluencies, he might be considered to be stuttering.
In summary, we can say that one measure that will help us distinguish a child with borderline stuttering from a normally disfluent child is a higher proportion of partword and monosyllabic whole-word repetitions and prolongations compared with multisyllabic word and phrase repetitions. In the next section, we see that children who show the types of tension in their disfluencies that have been labeled blocks, broken words, and dysrhythmic phonations are beginning rather than borderline stutterers.
The number of times a word or sound is repeated in a part-word or monosyllable word repetitive disfluency appears to be another sign that distinguishes children who stutter from their normally disfluent peers. In Yairi’s (1981) sample of 33 nonstuttering children, repetitions typically involved only one or two extra units of repetition (e.g., one extra unit would be li-like this). Other studies comparing stuttering and nonstuttering children (Ambrose & Yairi, 1995; Johnson & associates, 1959; Yairi & Lewis, 1984; Zebrowski, 1991) have found that the repetitive disfluencies of nonstuttering children average 1.13 extra units and that of stuttering children, 1.51. Thus, the frequent occurrence of repetitions having more than one extra unit is a warning sign of borderline stuttering.
We have said that borderline stuttering consists primarily of effortless repetitions and occasional prolongations. However, as Van Riper (1971, 1982) and Bloodstein (1995) note, these young children are often highly variable in their stuttering. Although they usually show the core behaviors of borderline stuttering, they may have brief periods of normal fluency as well as days when they show signs of more advanced stuttering.
Secondary Behaviors
A younger preschool child with borderline stuttering has few, if any secondary behaviors. The degree of tension may sometimes seem to be slightly greater than normal, but these children don’t seem to increase tension and struggle like older preschool children do. Children with borderline stuttering also do not exhibit accessory movements before, during, or after stutters. In fact, there is often nothing in their behavior to indicate that they are aware of their stutters. Some children with predominantly borderline stuttering may go through periods in which their stuttering suddenly escalates to the level of beginning stuttering, with tension and some other secondary behaviors, but then it falls back again to the borderline level.
Feelings and Attitudes
Because children with borderline stuttering seem to have little awareness of their stutters, they do not show concern or embarrassment. When they repeat a sound or a syllable, even five or six or more times, they usually go on talking as though nothing has happened. One exception, however, is that once in a while children with borderline stuttering might appear surprised or frustrated when they are repeating a syllable several times and are unable to finish a word. Then, they may stop and cry out, “Mommy, I can’t say that word,” or otherwise demonstrate brief alarm or surprise. But in general, these younger preschool children show little or no evidence of awareness that they have disfluencies that are different from those of their peers, and at this age, peers usually don’t react to the stuttering. Table 7.4 summarizes the major characteristics of the younger preschool children’s borderline stuttering.
Table 7.4 Characteristics of Borderline Stuttering in a Younger Preschool Child

Underlying Processes
I hypothesize that the symptoms of borderline stuttering result from the constitutional, developmental, and environmental factors described in Chapters 2, 3, 4, and 5. The constitutional factors associated with borderline stuttering often first show their effects as an excess of normal disfluencies. As previously stated, environmental and developmental pressures may be great between 2 and 3.5 years, and it is during this period that borderline stuttering typically emerges. The converging demands of expressive language and motor speech development ordinarily peak about this time “when an explosive growth in language ability outstrips a still-immature speech motor apparatus” (Andrews et al., 1983, p. 239). This age is also filled with psychosocial conflicts as a child copes with security needs as an infant while striving to become more independent as a toddler. The child may be ready to explore but is also fearful. The birth of a new brother or sister may trigger the child’s insecurity with the threat of being replaced. An older sibling may turn belligerent toward him because of the older child’s own need to express aggression as a prelude to puberty. Just as these stresses wax and wane in strength during preschool years, so does the child’s stuttering.
As children mature, certain developmental stresses may taper off. After age 5, children may feel more integrated within themselves and within their families.
Articulation and language skills, although still not at adult levels, have been mastered sufficiently for most children to say what’s on their minds and to be understood. They have also mastered other motor skills, such as walking and running, as well as riding a tricycle or a bike with training wheels. They may have adjusted to a new, younger sibling as well and made at least temporary peace with an older one.
By now, the capacities of many of the children who had modest predispositions to stutter can easily meet most environmental demands. Therefore, many of those who were borderline stutterers will have acquired normal fluency skills by the time they are 4 or 5 years old. Others may still have many disfluencies at this age, but will eventually outgrow them, perhaps because they are not frustrated by them and do not respond with tension or by rushing. In general, they are functioning well, feel accepted, and can use their resources to compensate for whatever difficulties in speaking remain.
Some children, of course, do not outgrow borderline stuttering. They may continue to stutter, and their symptoms may worsen. They may be children who have substantial predispositions to stutter, which cannot be offset by a “good enough” environment (Winnicott, 1971). Their ability to produce speech and language at the rate and level of complexity used by parents and peers may be insufficient. And their continuing efforts to meet advanced speech and language targets may result in excess disfluency that does not diminish as they pass their third and fourth birthdays. Their frustration tolerance for the repetitions that 2- and 3-year-olds exhibit may be low. Rather than shrugging them off, they may begin struggling to produce flawless speech, thereby placing greater demands on their speech production and emotional resources. Still other children may continue to stutter because environmental and developmental stresses do not diminish. Their insecurity may continue from sibling rivalry, breakup of the family, or a parent’s death. They may have language or articulation problems, as well as stuttering, which limit their communication abilities throughout their preschool years.
Deficits in the processes underlying speech and language development, plus the frustration of being unable to communicate easily, may be devastating to fluency. This may result in the increased tension we see in older preschool children with beginning stuttering. A child in this situation is unlikely to outgrow stuttering unless parents and professionals provide extensive support.

A video clip of a child with borderline stuttering is available on thePoint.
1. Younger preschool children with borderline stuttering usually exhibit a greater amount of disfluency than do normal children–more than seven disfluencies per 100 words.
2. Using another measure of frequency, the proportion of stuttering-like disfluenciesmay be greater than half of all disfluencies.
3. Children with borderline stuttering are also likely to repeat units more than once inmany of their part-word and monosyllabic word repetitions and to have many more part-word and monosyllabic word repetitions and prolongations than multisyllabic word and phrase repetitions, revisions, and interjections.
4. At the same time, their disfluencies, like those of nonstuttering children, are usuallyloose and relaxed appearing. Also, like nonstuttering children, children with borderline stuttering show little or no awareness of their speaking difficulty. Only rarely do they express frustration about it.
5. Among the underlying processes behind borderline stuttering are probably some ofthe speech and language-processing anomalies described in the earlier chapter on constitutional origins of stuttering. Such deficits in resources may interact with the demands of speech and language development, the pressure from higher rates of speech, more complex language, competitive speaking situations, and other attributes of a normal home. In addition, some of the psychosocial conflicts described earlier that increase normal disfluency are likely to be active in creating borderline stuttering.
In the older preschool child, (Fig. 7.4) stuttering usually has more tension and hurry than stuttering in a younger child. It may have evolved over a period of months or a year or two from borderline stuttering that the child manifested earlier. Or it may appear suddenly in an older preschool child during a time of stress or excitement. The tense and hurried stuttering may alternate with looser, easier disfluencies. Gradually, this more advanced type of stuttering will become commonplace. Soon, the child becomes impatient with his stuttering as it is happening–perhaps even embarrassed– and may begin to use a variety of escape behaviors as a consequence. For example, he may try to end long repetitions by using an eye blink or head nod. Periods of increased stuttering may last for several months, but periods of fluency may last only a few days. As these signs occur more consistently, tension increases and struggle is more evident. Instrumental and classical conditioning processes increase the frequency of struggle behaviors, complicate the child’s pattern of stuttering, and spread the symptoms to many more situations.

Figure 7.4 Child who may be a beginning stutterer.
As mentioned, some children exhibit beginning stuttering at onset, without passing through a stage of borderline stuttering. Van Riper (1971, 1982) described several different profiles of stuttering with tense blockages at onset. Many of the children he depicted as more severe at onset were relatively older (e.g., 4, 5, or 6 years old) when their stuttering first appeared. Onset in these children seemed to be related to one of two factors: delayed language development or emotional events. In a study of the onset of stuttering, Yairi and Ambrose (1992b) described onsets of stuttering that were characterized by the signs I described for beginning stuttering in 28 percent of their sample of 87 children. Many of these children had relatively sudden onsets, with normal fluency changing to beginning stuttering within one day or at most one week.
Core Behaviors
The core behaviors of beginning stuttering differ from those of borderline stuttering in several ways. Repetitions begin to sound rapid and irregular. The final segment of a repeated syllable often sounds abrupt; if it is a vowel, it will sound as if it were suddenly cut off or were a neutral or schwa vowel ( “uh”) that had been substituted for the appropriate one, as in “luh-luh-luh-like” instead of “li-li-li-like.” Repetitions are also produced more rapidly, sometimes with an irregular rhythm. Rather than patiently repeating a syllable as a borderline stutterer does, a child with beginning stuttering hurries through repetitive stutters, as though juggling a hot potato.
As symptoms progress, a beginning stutterer increases tension throughout the speech mechanism. Stuttering is sometimes accompanied by a rise in vocal pitch, resulting from increased tension in the vocal folds. Rising pitch may first appear toward the end of a string of repeated syllables, but over time will appear earlier in the repetitions. A child with beginning stuttering sometimes prolongs sounds that he would have previously repeated. Initially, he may prolong the first sounds of syllables, but as stuttering grows more severe, he may also prolong middle sounds, and they too may be accompanied by an increase in pitch.
As beginning stuttering progresses, the first signs of blockages appear. These are significant landmarks, which indicate that a child is stopping the flow of air or voice at one or more places (Van Riper, 1982). He may inappropriately jam his vocal folds closed or wide open, interrupting or possibly delaying the onset of phonation (Conture, 1990). Shutting off the airway is usually heard as a momentary stoppage of sound in a child’s speech and is sometimes accompanied by visual cues; the child may seem momentarily unable to move his mouth or may make groping movements with his mouth as he tries to get air or voice going again. When the stoppage of movement, voice, or airflow first begins, it may be so fleeting that we don’t notice it unless we are listening and watching carefully. As these blocks worsen, they become so obvious that they may overshadow the repetitions and prolongations that may remain.
Secondary Behaviors
As these older preschool children’s symptoms progress, secondary behaviors are added. They are called secondary because they appear to be responses to the runaway repetitions and increased muscle tension that have emerged. In addition, although hard evidence is lacking, the core behaviors of tension and speeding up seem to be “involuntary,” to have begun as a reaction that is beyond the stutterer’s ability to control. In contrast, the secondary behaviors we are describing seem to have begun “voluntarily.” They are–at least initially–deliberate.
Among the earliest of the secondary symptoms are “escape” behaviors, which are maneuvers used to end a stutter and finish a word. Children with beginning stuttering often show escape behaviors after several repetitions of a syllable. They may nod their heads, squint their eyes, or blink just as they try to push a word out. This extra effort often seems to help–in the short run. For the moment, they escape from the punishing repetition, prolongation, or block. Alternatively, they may insert a filler, such as “uh” or “um,” after a string of fruitless repetitions. The “um” seems to release the word, perhaps by relaxing the tightly squeezed larynx or by unlocking the lips. The “um” can always be said fluently, and once uttered, phonation and movement for the word often begin. The fillers work like a little push you might give your sled if it were stuck in the snow as you start down a hill; the “um” gets the child going again when he is stuck in a stutter.
A beginning stutterer starts to use escape behaviors earlier and earlier in stutters. The first appearance of these behaviors is usually after a child has repeated a sound quite a few times and is thoroughly frustrated about it. It may sound this way: “LuhLuh-Luh-Luh-Luh-umLet’s go!” Soon, however, the child will not wait until she has tried to say the sound five times. She finds herself about to say a word, feels convinced it won’t come out, and then perhaps instinctively uses escape behaviors when she is first starting to stutter: “L-umLet’s go!” Such “starters” may even appear before the first sound of the word, in this fashion: “umLet’s go!” This is really an avoidance behavior (because it is deployed to avoid a stutter before being stuck in one); they are more common among children with intermediate stuttering, even though it occasionally appears in the speech of a child with beginning stuttering.
Feelings and Attitudes
An older preschool child with beginning stuttering has stuttered many times. He is aware of stuttering when it happens. The feelings a beginning stutterer has just before, during, and after stutters are often strong. Frequently, frustration is a major feeling. The child may stop in the middle of a stutter and say, “Mom, why can’t I talk?” However, such momentary frustration grows into momentary fear when a word or sound is stuck for several seconds, and the child feels helpless and out of control.
Although a child with beginning stuttering is conscious that he has some “trouble” when he talks, he has not yet developed a belief that he is a defective speaker. This lack of a negative self-image may be attributed, as Bloodstein (1987) and Van Riper (1982) have suggested, to the “episodic” nature of stuttering. Sometimes it’s there; sometimes it’s not. Sometimes a child feels that he has problems when he talks; other times he forgets about it. The essential characteristics of beginning stutterers are presented in Table 7.5.
Table 7.5 Characteristics of Beginning Stuttering in an Older Preschool Child

Underlying Processes
The signs and symptoms of beginning stuttering in the older preschool child can be observed by any experienced clinician. We have seen them in hundreds of children who stutter. But the processes underlying these behaviors are not so easy to see. In Chapters 4 and 5, we suggested that beginning stuttering may result from the interplay between constitutional and environmental factors, especially in a child with a reactive temperament. I will review my hypotheses about the core behaviors of beginning stuttering as well as the learning processes that are likely to perpetuate the core behaviors and a child’s secondary reactions.
Increases in Muscle Tension and Tempo
One of the first signs of beginning stuttering in older preschool children is the appearance of extra muscular tension in repetitions and prolongations and increased tempo or rate in repetitive stutters (Van Riper, 1982). Why do these changes occur? Oliver Bloodstein (Bloodstein, 1987; Bloodstein & Ratner, 2008) suggests that facial tension and strained glottal attacks in the speech of young children who stutter may reflect the extra muscular effort that emerges when they anticipate difficulty. Edward Conture (1990) offers a related view. He sees the increased articulatory and laryngeal muscle tension as a child’s attempt to control sound-syllable repetitions, which are so distressing to him and his listeners. We have described such tension as a child’s effort to control a frustrating and scary behavior of his own body, an attempt to stiffen the speech muscles and brace himself against the perturbations of seemingly involuntary, runaway repetitions (Guitar, Guitar, Neilson, O’Dwyer, & Andrews, 1988). One can imagine this taking place in the same way that a child who is attempting to ice skate may begin to stiffen in response to rough spots in the ice by stiffening and assuming a less than ideal stance for continued forward movement.
The other early sign of beginning stuttering–increases in the rate of repetitive stutters–is cited by a number of authors as an indication that stuttering is worsening. Van Riper (1982), in describing the developmental course of the majority of children whose stuttering persists, stated that “the tempo changes as the disorder develops. The repetitive syllables become irregular and are often spoken more rapidly than other fluent syllables.” Starkweather (1987) explained this increase in the speed of repetitions as a product of the pressure that children feel as they become more aware of the extra time it takes them to produce an utterance.
But why are these increases in tension and tempo so common in the development of stuttering, and why are they so difficult to change in therapy? In Chapter 6, I described my view that children in whom stuttering persists may be especially sensitive to certain kinds of experiences. Faced with frustration or fear, I hypothesized, they would react with elements of their biologically based freezing or flight responses. The signs of beginning stuttering appear to have similarities with such reactions. The excessive muscle tension in stuttering typical of older preschool children can be viewed as a way that the limbic system can cause children to freeze in the face of their frustrating or frightening repetitive disfluencies, transforming them into abrupt, tense repetitions, blocks, or prolongations. As I have mentioned, some children appear to show tense blocks at the onset of their stuttering. These may be children who have high degrees of emotional sensitivity and whose very first manifestation of stuttering may result from fear-based responses to speaking experiences.
Research bears out the speculation that at least some adults who stutter contract their muscles in such a way that movement and phonation are immobilized. Freeman and Ushijima’s (1978) and Shapiro and DeCicco’s (1982) studies indicate that stuttering is associated with abnormal muscle co-contraction of adductor and abductor muscles in the larynx. Such co-contraction could produce stiffening of the phonatory structures and silencing of vocal output. Other studies of stuttering have demonstrated co-contraction in articulatory structures (Fibiger, 1971; Guitar et al., 1988; Platt & Basili, 1973), which could also produce immobility and silence.
Unfortunately, little research directly supports the notion that the increased rate of repetitions reflects the flight response. We have some preliminary evidence that stutterers have more rapid productions during repetitions than do nonstutterers. An unpublished study (Allen, 1988) carried out in our clinical laboratory indicated that the durations of beginning stutterers’ repeated segments and the silences between them were shorter than the durations in similar disfluencies of nonstuttering children matched for age. This finding has been confirmed in the work of Throneburg and Yairi (1994), who also found that the silent intervals and the total durations of repetition disfluencies were significantly shorter in stuttering children compared with those of nonstuttering children controls. Such shortening of segments results in a faster speech rate, at least for the stuttered elements, and may reflect the “great increase in activity” seen in the flight response, although these particular data do not exclude the possibility that stuttering children were more rapid speakers to begin with. It may be relevant at this point to note that Kloth, Janssen, Kraaimaat, and Brutten (1995) found that rapid speaking rate was a predictor of which young children who were fluent at the time of testing but had family histories of stuttering would eventually stutter. The rapid rate in these children might be related to a reactive limbic system, although there is no evidence that speech rate is related to such reactivity.
The possibility that increased muscle tension and rapid repetitions are a result of biologically based freezing or flight responses is highly speculative at this time. If these responses are part of humans’ neural wiring designed for survival, this may be a potential explanation of why some children develop stuttering so rapidly and why tension responses are so difficult to change.
Effects of Learning on Stuttering
It is not clear whether the increases in tension and rate are voluntary or involuntary or both, but I have observed repeatedly that these reactions increase with persistent stuttering. This is likely to be the result of classical and instrumental conditioning, which are thought to be the forces at work when stuttering escalates from an occasional tense and hurried repetition to speech that is riddled with tense and hurried repetitions. Learning processes also turn occasional eye blinks or head nods during stutters into stereotyped patterns of blinking and nodding that transform innocuous disfluencies into something so obvious that listeners gape in surprise and parents look away. Let us examine these learning processes more closely so that we can better understand those who come to us for help.
Classical conditioning, I surmise, is responsible for previously “neutral” experiences and situations becoming able to elicit tense and hurried stuttering in a child’s speech. This occurs after many experiences in which a child’s repetitive stutters have brought on emotions that trigger increases in tension and hurry. Through the repeated pairing of the neutral situations with negative emotions related to stuttering, classical conditioning spreads this change in stuttering behaviors to more and more situations. In few children, tension appears at the onset of the disorder because these children experience a high degree of emotion during a speaking situation. Many children whose stuttering persists may be highly susceptible to classical conditioning because their reactive nervous systems are quick to produce emotions.
If classical conditioning spreads negative emotion to more and more speaking situations, it is instrumental conditioning that is responsible for the increase in frequency of escape behaviors in older preschool children’s beginning stuttering. This is because the child is rewarded for such things as head nods or eye blinks when they are followed by gratifying releases of the words on which he is stuttering. Instrumental conditioning generalizes escape behaviors to more and more situations and causes escape behaviors to occur earlier and earlier in stutters so that escape behaviors may eventually become “starters.”

A video of beginning stuttering is available on thePoint.
The principal differences between borderline stuttering seen in younger preschool children and beginning stuttering common in older preschool children are these:
1. The older child with beginning stuttering shows more tension and “hurry” in hisstuttering. This is often manifested in abruptly ended syllable repetitions, irregular rhythms of repetitions, evident stoppages of phonation, and momentarily fixated articulatory postures. Older preschool children also evidence such secondary behaviors as escape devices and starters. In addition, these beginning stutterers see themselves as persons who have trouble talking.
2. A major factor underlying beginning stuttering appears to be a child’s sensitivity tostress, which may result in frustration, triggering tension responses.
3. Classical conditioning then links such unconditioned response sensitivity to disfluency. When the child is disfluent, he feels threatened, frustrated, or afraid, and this in turn leads to the rapid, tense disfluencies that begin to appear in beginning stutterers. After repeated pairings, disfluency itself, rather than the emotion, elicits increased tension and rate. Classical conditioning also links a child’s disfluency to more and more people and places.
4. A third factor in beginning stuttering in older preschool children is instrumentalconditioning, which increases and then maintains the use of escape devices. These behaviors are negatively reinforced when a stutterers’ frustration is terminated by an escape behavior and are positively reinforced when the stutterer completes his communication.
The school-age child with intermediate stuttering (Fig. 7.5), who is typically between ages 6 and 13 years, has two major characteristics that distinguish him from a child with beginning stuttering. First, he is starting to fear stuttering, whereas older preschool children with beginning stuttering are usually only frustrated, surprised, or annoyed by it. Second, the school-age child with intermediate stuttering reacts to his fear of stuttering by appearing to avoid it, something beginning stutterers don’t do with any regularity. These new symptoms emerge gradually as a young stutterer experiences negative emotion more frequently during stuttering. For example, when the school-age child blocks and feels helpless, listeners respond with discomfort, pity, and ridicule. After this has happened frequently, he becomes afraid.

Figure 7.5 Child who may be an intermediate stutterer.
This fear may be attached first to the sounds and words on which he stutters most, and he starts to believe that these sounds are harder for him. Then he begins to scan ahead to see whether he might have to say them. When he anticipates that he will, he tries to avoid them. For example, he may say, “I don’t know,” to questions or substitute “my sister” for his sister’s name when talking about her. Sometimes, he may start a sentence, realize a feared word is coming up, then switch the sentence around to avoid stuttering, and end up producing a maze of half-finished sentences. With tactful questioning, the clinician can verify these avoidances.
A school-age intermediate stutterer’s fear of stuttering may be associated with situations as well as words. The youngster may find that he stutters more in some situations than in others. At first, he approaches these situations with dread, but later, he may go to great lengths to avoid them. Van Riper (1982) suggested that the development of such situational fears and avoidances depends on listener reactions. Consequently, counseling or advice for key listeners in a stutterer’s environment may help prevent them.
Core Behaviors
What are intermediate stutterers’ moments of stuttering like when they don’t avoid them? What are the core behaviors? Although they still repeat and prolong, their most notable core behaviors are now blocks. The blocks of children with intermediate stuttering seem to grow out of the increasing tension seen initially in beginning stuttering. A child at the intermediate level usually stutters by stopping airflow, voicing, movement, or all three, then struggling to get his speech going again. His stutters seem to surprise him less than when he was a beginning stutterer. Instead, as evidenced by his voice and manner in certain situations, he anticipates stutters.
I have the impression that the intermediate stutterer’s blocks are frequently characterized by excessive laryngeal tension, but he often blocks elsewhere, as well. He may squeeze his lips together, jam his tongue against the roof of his mouth, or hold his breath. Even though he is not highly conscious of just what he’s doing during a block, he has a vivid awareness that he is stuck, that he feels helpless, and that the word he wants to say won’t seem to come out.
A school-age child described his feeling of being blocked as like “a rock stuck in my throat.” When he was lucky, he said, a little army of men would come into his throat and break the rock into little pieces, breaking the block so that sounds would come out. He was describing the experience of first being totally stuck, then rapidly repeating the first segment of the sound as he fought his way out of the block. A common example is the “…uh-uh-uh-I” that you will hear when someone is blocked on “I” and tries to push through it. At first, there is a moment of silence and then the rapid, staccato first segment of the sound as the stutterer gets his larynx vibrating while maintaining a static articulatory posture. The larynx is still very tense; vibration stops and starts again and again. The vowel–either at the beginning or end of the syllable–is often what Van Riper (1982) called the “schwa” or neutral vowel. In fact, it is only the first, brief segment of the intended vowel, which is cut off too abruptly to be perceived as the sound normally used in the word. Inexperienced clinicians sometimes mistakenly categorize these repeated parts of blocks as repetitions, not realizing the stuttering has advanced from repetitions to blocks.
In addition to repetitions of parts of sounds, blocks can have prolongations in them. Sometimes, as he is pushing through a block, a stutterer will momentarily prolong a continuant sound as in “…mmm…mmmm…my.” Again, this probably results from the stutterer’s larynx vibrating momentarily, then seizing up again, and then vibrating again. I catalog such events as blocks rather than prolongations, because I think the core behavior is a complete stoppage of speech, even though it is mixed with momentary releases of laryngeal vibration. This confusing situation probably results from the fact that the sequence of repetitions, prolongations, and blocks reflects basically similar behaviors along a continuum of increasing tension, particularly in the larynx, as stuttering progresses.
Secondary Behaviors
The blocks just described can be devastating to a child who stutters. He is frustrated not only with his inability to make a sound, but he is often faced with surprised and uncomfortable listeners as well. Even patient listeners may not know what to do. They may interrupt, look away, or fidget, leaving the child or adolescent to conclude that he is doing something wrong and should try to escape or avoid these painful moments.
The escape behaviors that a speaker uses to free himself from stutters are present in preschool children with beginning stuttering, but they occur far more frequently in school-age children with intermediate stuttering. They are often more complex, too. An intermediate stutterer may blink his eyes and nod his head in an effort to escape a block. Sometimes, he may do both, and if he is still unable to say the word, he may resort to yet another device, such as slapping his leg. As these patterns grow more complex, they may also become disguised to look like natural movements and are performed more rapidly.
In addition to escape behaviors, a child at the intermediate level develops both word and situation avoidances, as previously mentioned. Word avoidances appear after he has had repeated difficulty with a particular word or sound and has discovered how to take evasive action before he has to say it. For example, a young stutterer in our clinic had been asked his name by a particularly stern teacher. He blocked severely on it and subsequently became fearful of saying his name, as well as other words starting with the same sound. He could usually think up synonyms for other words but found it awkward to substitute anything different for his name. So he learned to get a running start in saying his name by beginning with “My name is…” whenever he was asked his name. This permitted him to avoid stuttering about half of the time. It is a subtle form of avoidance that many clinicians call “starters.” More obvious examples of avoidances are given in the following paragraph.
Van Riper’s (1982) catalog of word avoidance techniques included starters (beginning a word by saying another word or sound, such as “well” or “uh” just before saying it); substitutions (substituting a word or phrase for another when stuttering is expected, as in “he’s my unc-unc-unc … my father’s brother”); circumlocutions (talking all around a word or phrase when anticipating stuttering, as in “well, I went to … yes, I really had a good time there, I saw the Empire State Building”); postponements (waiting a few beats or putting in filler words before starting a word on which stuttering is expected, as in “My name is………. Bill”); and antiexpectancy devices (using an odd manner or funny voice to avoid stuttering when it’s anticipated). I had a client in Australia who could only tell jokes fluently if he put on an accent that sounded like he came from Mississippi or Alabama.
Like escape behaviors, word avoidance techniques often become more rapid and more subtle with time. Indeed, some stutterers can disguise word avoidances to look like normal behavior. For example, they may put on pensive facial expressions and appear to search for a word while postponing their attempt to say a feared sound. Experienced clinicians learn to pick up subtle cues in the rate and manner of speaking that tip them off to the use of such avoidances. Such avoidances can be explored by the clinician and client at the appropriate moment in treatment.
Situational fears and avoidances also begin to appear in the school-age intermediate stutterer. Past stuttering in specific places or with specific people are the seeds from which situational fears grow. In school, stutterers usually have trouble reading aloud or giving oral reports. Most people who stutter, and even many nonstutterers, dread those classes in which teachers call on students by going up and down the rows. As in an earlier example, the students’ fears steadily mount as a teacher goes down the row, getting closer and closer to calling on them. Then, if called on, they may take a failing grade rather than give the oral report. In contrast, other school situations, especially casual ones like gym class or lunch period, are likely to hold little fear or expectation of stuttering for them.
Situational fears quickly generate situation avoidances. The student who fears giving answers in class may try to slouch low in his seat in hopes of being overlooked.
A stutterer who is afraid of making introductions will contrive ways of having other people make them. In junior high school, I coped with my fear of ordering in restaurants by ducking into the bathroom when the waitress approached our table, leaving my friends to order a cheeseburger for me. Every stutterer has his own pattern of situation avoidances, which may provide an important focus for therapy in many cases.
Feelings and Attitudes
Students with intermediate stuttering have gone well beyond the momentary frustration and mild embarrassment experienced by those with beginning stuttering. They have felt the helplessness of being caught in many blocks and runaway repetitions. The anticipation of stuttering and subsequent listener penalties have been fulfilled many times. These experiences pile up like cars in a demolition derby to create an entanglement of fear, embarrassment, and shame that accompanies stuttering. These feelings may not be pervasive or dog a stutterer all the time. However, stuttering has now changed from an annoyance to a serious problem.
A major influence on such students’ feelings is the cognitive development, begun at age 3 or 4, that enables him to compare himself with his peers. Once he begins school, peers have a greater and greater influence on him. He may stutter more as he encounters new people and new situations, and as he does, peers may begin to ask him why he talks the way he does and to make comments about his stuttering or tease him about it. As a result, increasingly negative self-awareness about his speech leads to feelings of embarrassment, shame, and guilt.
A student with intermediate stuttering shows his increasingly negative feelings about stuttering in many ways. He may look away from listeners when he is stuttering and flush with embarrassment immediately afterward. He may become stiff and uneasy at the prospect of speaking. His stuttering pattern includes an increasing number of avoidance devices, and he is beginning to evade situations in which he feels he may stutter. These are all signs that his feelings and attitudes are becoming suffused with fear. Table 7.6 gives the characteristics of intermediate stutterers.
Table 7.6 Characteristics of Intermediate Stuttering in a School-Age Child

with hope as treatment begins. Figure 7.6 was drawn by a young man in his first few weeks of therapy. It reflects his extensive negative feelings on the left side; he has drawn himself in a jail cell with tears/rain falling around him and the key just out of reach. On the right side of the drawing, he shows himself after therapy, escaping from jail, running in the sunshine with grass underfoot and a flower in the background.

Figure 7.6 A young man’s drawing of himself as he begins therapy (left side) and his hopes for a happy outcome (right side). Drawn by Marcel Etienne.
Underlying Processes
Many of a school-age, intermediate stutterer’s symptoms result from the same processes that underlie those of beginning stutterers. There are major differences, however. In intermediate stuttering, classically conditioned tension responses are more evident, conditioned frustration is now becoming a more intense fear reaction, and avoidance conditioning has become a factor in shaping stuttering behaviors.
Avoidance conditioning transforms escape behaviors, such as the use of “um” to escape from a stuttering block, into avoidances, such as saying “um” before saying a word on which stuttering is expected. This learning process also leads students with intermediate stuttering to avoid words, to change sentences around, and to avoid speaking situations entirely. Avoidance learning also generalizes from one word to another and from one situation to another.
Avoidance conditioning may proceed very quickly in people with persistent stuttering because they may have a genetic or congenital bias toward right-hemisphere, emotionally based behaviors, as we described in Chapters 2 through 6. The threat of stuttering may elicit “prepared” defensive reactions, such as avoidances of words or situations. Such avoidances are strongly maintained because individuals who have developed them use them when they anticipate stuttering, which decreases or eliminates the fear. Thus, avoidances are maintained by negative reinforcement. By avoiding the stuttering, individuals who stutter never have the opportunity to discover that stuttering is not so painful after all. Therapy must (a) structure situations to help them learn this and (b) give them new behaviors to substitute for the old avoidances.

A video of intermediate stuttering is available on thePoint.
Intermediate stuttering in school-age children is differentiated from beginning stuttering in older preschool children by the following:
1. There are increasingly tense blocks, repetitions, and prolongations; the increasedtension results from feelings of frustration, fear, and helplessness. These feelings trigger tension responses, which interfere with fluency and in turn produce more frustration, fear, and feelings of helplessness. As tension mounts, this vicious cycle continues; blocks are longer and more noticeable, more listeners react with surprise and impatience, and the student’s fear increases in response to these reactions.
2. The increasing presence of fear and anticipation of bad experiences spurs the student to develop avoidance behaviors in addition to the escape behaviors he is already using. Avoidance conditioning is difficult to undo.
3. The child with intermediate stuttering increasingly feels embarrassment, shame,and guilt as he realizes that his speech is markedly different from that of his peers.
Individuals whose stuttering has persisted into older adolescence and adulthood (Fig. 7.7) typically have a deeply ingrained pattern of core and secondary behaviors. Often stuttering is a major player in their school, work, and social lives. They may avoid talking in class, decline job opportunities, and limit their social activities from fear of stuttering. This describes me at age 20.

Figure 7.7 Individual who may be an advanced stutterer.
Of course some older teens and adults stutter only mildly or aren’t bothered by their stuttering; they carry on their lives seeing it as a minor annoyance. These individuals often don’t seek treatment–unless their stuttering suddenly gets in the way of something they want to do. One of my clients who had relatively mild stuttering was in the U.S. Air Force and wanted to move up from navigator to pilot. This was during the Vietnam War, and I worried that this promotion would put him more at risk. Nevertheless, we worked hard together for six months, and he made the grade.
Treatment of older teens and adults differs from treatment of younger stutterers because the client can take much of the responsibility for therapy including substantial work outside the clinic.
An advanced stutterer’s increased capacity for independent work may compensate for another characteristic of this level–a long history of stuttering. Patterns of stuttering with tension, escape, and avoidance behaviors are now firmly established. Emotions such as frustration, fear, guilt, and hostility have built up over many years of being unable to speak like other people and many bad experiences with thoughtless, uninformed, or momentarily startled listeners. Beliefs are usually distorted by the conviction that other people are impatient or disgusted by the speaker’s stuttering.
After many years of stuttering, adults and adolescents who stutter increasingly think of themselves as stutterers rather than as people who have occasional difficulty speaking. Except for a few safe situations in which they may be relatively fluent, they have some fear for most speaking situations, and they shape their lives accordingly. They may believe that their stuttering is as noticeable to others as though they had two heads–and nearly as unacceptable.
Core Behaviors
Core behaviors of stuttering in older teens and adults include repetitions and prolongations, but advanced stuttering is often distinct in the struggle and tension of blocks–stoppages of sound and movement. Advanced stutterers may block, then release a little sound only to fall back into the block again. It might sound like this: “(silence)…m-m-m…(silence)…m-m-muh…(silence)…my (said with a sudden effort)…name is Barry.” Such behaviors may be longer and more struggled in clients with advanced stuttering than in school-age kids who have intermediate stuttering, but they are essentially the same. Blocks may be associated with tremors. During blocks, tremors of the lips, jaw, or tongue may be apparent. As you may remember from the discussion of tremors in Chapters 2, 3, and 6, tremors appear in those who have been stuttering for several years and may occur when stuttering is accompanied by strong emotion.
In a few advanced stutterers, blocks are hardly evident at all. They may have honed their avoidance devices to such a fine edge that core behaviors are scarcely noticeable. If stuttering does become evident, it usually feels devastating. Consequently, much of their energy is spent anticipating blocks that often don’t occur and mustering avoidances to keep anxiety at bay. One such individual, a delightful woman I knew and whom I’ll call Lenore, said she had stuttered since childhood. Yet, she almost never had a repetition, prolongation, or block that I could see. She was highly competent at everything she did, but severely limited her life because of her fear that she would stutter. In particular, she often felt she came across as far less articulate than she might have because of the frequency with which she substituted words to avoid stuttering.
Older teens and adults with advanced stuttering, like school-age youngsters at the intermediate level, have repetitions as well as blocks. These are not the easy, regular repetitions of borderline stuttering, but are more like those of beginning stuttering– tense, with a rapid, irregular tempo. They may be repetitions of syllables, luh-luh-luhlike this, or mixed with fixed articulatory postures of tense blocks, l …l … luh-luh-luh … like this. The latter look as if the speaker recoils from a momentary fixation then gets stuck again.
Secondary Behaviors
Advanced stuttering in older teens and adults involves many of the same word and situational avoidances that are seen in intermediate stuttering, but the avoidances are likely to be more extensive. Some behaviors are more obvious than others. When I was in high school, I used several avoidance devices that often didn’t work, such as
“uh … well … you see” and a gasp of air, followed by a block of long duration filled with unsuccessful escape attempts before I finally released the blocked word with great effort. Other advanced stutterers may approach feared words cautiously and use subtle mannerisms, such as appearing to think just before saying them, so that most listeners don’t realize they are stuttering. These stutterers are usually on guard much of the time, scanning ahead with their verbal early-warning systems.
Many advanced stutterers also control their environments carefully so that they can avoid situations in which they are likely to stutter. They may feign sickness when they have to give a speech, use answering machines rather than answering the telephone, or arrange to have their spouses or children deal with store clerks. Often, with careful questioning of advanced stutterers who use avoidances a great deal, you can learn what occurs when avoidances don’t work. Even the most skillful avoiders are sometimes caught with their defenses down and become stuck in a block. Core behaviors may also be elicited by asking some stutterers to stutter openly without using secondary behaviors. Stutterers who can do this, especially those who can do it without excessive discomfort, are more amenable to change.
Feelings and Attitudes
The feelings and attitudes of older teens and adults, like their stuttering patterns, have been shaped by years of conditioning. Over and over, they have learned that much of their stuttering is unpredictable. When it is predictable, it comes when they want it least–when they want more than anything to be fluent. As a result, they often feel out of control. Figure 7.8 reflects one individual’s depictions of his own feelings of being out of control when stuttering.

Figure 7.8 “How I feel when I stutter” by Mike Peace. (Courtesy of Dr. Trudy Stewart.)
These uncomfortable feelings are often buttressed by a stutterer’s perceptions of how others see him. Listeners’ reactions look overwhelmingly negative to him. Even when listeners say nothing, their faces appear to say everything. It is as though stuttering is a rattletrap car that always stalls in heavy traffic amid honking drivers. Such experiences gradually shape advanced stutterers’ attitudes toward feelings of helplessness, frustration, anger, and hopelessness.
Of course, individuals’ responses to stuttering vary greatly. If a person who stutters has many talents and abilities for which he is recognized and if he has an assertive personality, he may be less devastated by stuttering. The former CEO of General
Electric company, Jack Welsh, is a good example. But if the individual has a highly sensitive nature, his feelings and attitudes about stuttering may be an important component of his problem. The movie “The King’s Speech” suggested that Bertie, who was to become the King of England, George VI, was a sensitive soul who was debilitated by his stuttering until he received some very confidence-building treatment from Lionel Logue, his unorthodox Australian clinician.
The point is that by the time a stutterer is an adult, he has had years of experiencing stuttering, feeling frustrated and helpless, and has developed techniques to minimize pain. Unless he has strong attributes to compensate, he is likely to feel that stuttering is a big part of who he is to other people. It is a part that he hates, a part on which he blames many other troubles, and a part he wants to eliminate.
Some stutterers, however, who reach the advanced level have become reconciled to their stuttering. If they are in their 20s, 30s, or beyond, there may be some natural resistance to treatment, because stuttering has become part of their identities. After years of doubt and turmoil, they’ve grown accustomed to themselves as stutterers. To consider treatment is to reject a part of themselves, to open old wounds. Those who risk change, enter treatment, and succeed will find the risk to have been worthwhile. But those who enter treatment and do not succeed may suffer twice from the pain of failure as well as the loss of the denial or reluctant acceptance of stuttering that had been in place before the attempt at treatment but was given up.
Table 7.7 lists the major characteristics of advanced stutterers.
Table 7.7 Characteristics of Advanced Stuttering in Older Teens and Adults

Underlying Processes
Advanced stuttering, unlike lower levels of stuttering, is influenced less by its original constitutional, developmental, and environmental factors than by the older teen or adult’s reactions to his stuttering. The effects of home environments, developmental pressures of speech and language, and maybe even some differences in central nervous system function have been diminished by maturation and learning. However, conditioned habits that were learned in response to these early factors are stronger than ever. Their effects have been magnified by years of experience, and the way the brain operates in speech has been modified as a consequence. Moreover, an individual’s characteristic patterns of tension, escape behaviors, and word and situation avoidances have become almost automatic through years of practice. For example, he may exhibit a string of avoidance and escape behaviors but only remember that “the word got stuck.”
The older teen and adult’s stuttering is affected by higher-level explicit learning as well. He has developed a self-concept as an impaired speaker, which carries highly negative connotations for most. Self-concepts begin to be formed during preschool years and are based initially on what one can do, rather than what one is (ClarkeStewart & Friedman, 1987). More enduring traits are added as a result of social interactions in later childhood, adolescence, and beyond (Roessler & Bolton, 1978). Thus, a stutterer’s self-concept at the earliest levels of development is determined, in part, by his perception of how he talks. In a child’s early years, his impression of stuttering may be a fleeting awareness that he sometimes has difficulty talking. At later levels of development, the reactions of significant listeners–parents, peer group, other adults–have a major impact. Now his self-concept may become filled with relatively enduring negative perceptions as a result of listeners’ impatience and rejection. A negative self-concept is formed not only by perceptions of listeners’ reactions, but it in turn also affects those perceptions.
Researchers studying the psychology of disability suggest that “one’s perception of self influences one’s perception of others’ views of oneself, rendering social interaction more difficult” (Roessler & Bolton, 1978). Applied to clients with advanced stuttering, this suggests that they are likely to project their own rejections of stuttering onto listeners, thereby inhibiting interactions with them. This vicious cycle can only be stopped when an outsider helps a stutterer test the reality of his perceptions.
In addition to working on cognitive aspects of the problem, therapy for advanced stuttering also deals directly with the avoidances such clients have learned so well. As mentioned in the discussion of intermediate stuttering, as avoidance conditioning progresses, individuals fear not only words and situations, but also stuttering itself. By deconditioning this fear and changing such responses, treatment enables stutterers to stutter with less fear by associating the clinician’s approval with a calmer, more relaxed way of stuttering. Gradually, tension and hurry fade from disfluencies, they feel more in control, and their fears diminish even further as a result.

A video of advanced stuttering is available on thePoint.
The diagnosis of advanced stuttering in older teens and adults describes a developmental level and implies a particular treatment orientation as characterized by the following:
1. Treatment may be easier because the client can assume much of the responsibilityfor generalization beyond the clinic.
2. On the other hand, treatment is more challenging because the client with advancedstuttering has habituated patterns of behavior more deeply than at earlier levels. The advanced stutterer’s core behaviors often consist of long blocks with considerable tension and at times visible tremors. Secondary behaviors may consist of long chains of word avoidance and escape behaviors. Situational avoidance is common.
3. Some older teens and adult stutterers may hide and disguise their stuttering wellenough to avoid detection by many listeners, but this is at the cost of constant vigilance.
4. Feelings of frustration and helplessness usually accumulate over the years, leadingto coping behaviors and a lifestyle that may be highly constraining. Such responses create a self-concept of an inept speaker whose stuttering is unacceptable to listeners. This in turn affects the stutterer’s perceptions of the listener’s reactions.

• Table 7.8 summarizes the characteristics of the five developmental/treatment levels described in this chapter.
• Each individual who stutters will have his own course of development, influencedby the interaction of constitutional and environmental factors.
• The clinician needs to use her understanding of the underlying processes to designprocedures to treat each individual’s core behaviors, secondary behaviors, and feelings and attitudes.
Table 7.8 Characteristics of Five Developmental/Treatment Levels

1. In the “exceptions and variations” section of the overview, different types ofstuttering onset and development are described. What factors might cause these differences?
2. In discussing normal disfluency, it is suggested that “if a child shows what appears to be normal disfluencies, such as single-word repetitions, but consistently displays pauses or interjections of ‘uh’ immediately before or during disfluencies, he should be carefully evaluated as possibly stuttering.” What might be going on? What might these pauses or interjections signify?
3. The idea of a dyssynchrony in the timing of the elements of spoken languageproduction is suggested as an underlying process of normal disfluency. It is also used to account for primary stuttering. How can both types of disfluency be accounted for by the same process?
4. What is the difference between core behaviors and secondary behaviors?
5. At what ages is normal disfluency likely to be most frequent?
6. Name three influences that may cause normal disfluency to increase.
7. What are three ways in which core behaviors of normal disfluency differ fromthose of borderline stuttering?
8. Describe the core behaviors of the beginning stutterer.
9. What causes greater muscle tension in beginning stuttering compared to borderline stuttering?
10. Describe why an escape behavior is used by a stutterer. Give examples.
11. What is the major secondary behavior that differentiates the intermediate fromthe beginning stutterer?
12. Compare the feelings and attitudes of the borderline, beginning, and intermediatestutterers.
13. Describe the role of the listener in the development of the advanced stutterer’sself-concept.
1. Visit thePoint and download the video clips of speakers who are representative of each level of stuttering (normal disfluency, borderline, beginning, intermediate, and advanced), and play them in random order for your class. See how many of your fellow students can correctly identify each level.
2. Make audio or video recordings of a number of nonstuttering students in aclass and determine which of them are more disfluent and which are less disfluent. Is there a gradual continuum between more disfluent and less disfluent, or are there two distinct groups? Are any of the more “normally disfluent” students actually borderline stutterers? Should the term “borderline stutterer” be used only for preschoolers?
3. Read Yairi and Ambrose’s (2005) chapter on the development of stuttering(see Suggested Readings) and compare that perspective with the view presented in this chapter.

Bloodstein, O., & Ratner, N. (2008). Symptomatology. In A Handbook on Stuttering. San Diego: Singular Publishing Group, Inc.
The subsection titled, “Developmental Changes in Stuttering” in this chapter describes four stages similar to our levels of stuttering development. Other schemas of developmental changes are also discussed in a clear and logical style.
Gray, J. A. (1987). The psychology of fear and stress. Cambridge: Cambridge University Press.
This is a very readable exposition of relatively recent findings about innate fears, conditioning, and brain processes involved with escape and avoidance learning. Gray also describes his concept of the “behavioral inhibition system,” a model of the role of conditioning, language, the limbic system, and anxiety on behavior.
Luper, H. L., & Mulder, R. L. (1964). Stuttering: Therapy for children.
Englewood Cliffs, NJ: Prentice-Hall.
An excellent treatment text that describes four developmental levels of stuttering similar to the levels described here. Although out of print, this book is available for under $10 at
Starkweather, C. W. (1983). Speech and language: Principles and processes of behavior change. Englewood Cliffs, NJ: Prentice-Hall.
This book describes the principles of instrumental, classical, and avoidance conditioning that underlie much of stuttering behavior. It gives a clear account of how these principles create stuttering behavior and how conditioning is used in treatment.
Van Riper, C. (1982). The development of stuttering. In The Nature of Stuttering. Englewood Cliffs, NJ: Prentice-Hall.
In this chapter, Van Riper describes four developmental tracks of stuttering, three of which depart substantially from our stages of stuttering development. This chapter gives the reader a good sense of individual variability in stuttering.
Williams, D. F. (2006) Stuttering recovery: Personal and empirical perspectives. Mahwah, NJ: Lawrence Erlbaum Associates.
This book is an informal compendium of essays about the experience of stuttering, information about stuttering, and personal anecdotes about the things that happen to you when you stutter.
Yairi, E. & Ambrose, N. G. (2005). The development of stuttering. In Early Childhood Stuttering (pp. 141–195). Austin, TX: Pro-Ed.
This chapter reviews other authors’ descriptions of the development of stuttering and presents a different perspective on the changes that occur in stuttering from onset to recovery or persistence. The data provided support the view that 75 to 85 percent of children who begin to stutter will recover without treatment and that stuttering typically decreases in severity and frequency after onset.

Preliminaries to Assessment

The Client’s Needs
The Client’s Right to Privacy
Cultural Considerations
The Clinician’s Expertise
Assessment of Stuttering Behavior
Speech Sample
Assessing Frequency
Assessing Types of Stutters
Assessing Duration
Assessing Secondary Behaviors
Assessing Severity
Assessing Speech Naturalness
Assessing Speaking and Reading Rate
Assessing Feelings and Attitudes
Assessment of Preschool Children
Assessment of School-Age Children
Assessment of Adolescents and Adults
Continuing Assessment
After studying this chapter, readers should be able to:
• Understand how to discern the client’s needs and plan treatment around them
• Describe how to protect the client’s right to privacy and how awareness of theclient’s right to privacy can facilitate trust
• Explain why multicultural awareness is so important when working with clientsfrom different cultural and linguistic backgrounds
• Describe how a clinician can demonstrate her expertise about stuttering in away that will engender trust and motivation
• Explain why reliability in a measurement procedure is important and how reliability may be assessed
• Discuss the need for obtaining appropriate speech samples when assessing stuttering
• Explain the advantages and disadvantages of assessing frequency of stutteringand how frequency can be assessed effectively
• Explain why it can be useful to assess different types of stutters that a clientmay have
• Describe how duration of stutters may be important and how this can be assessed
• Discuss assessment of secondary stuttering behaviors
• Describe four tools to assess stuttering severity and explain when each might beused
• Explain why speech naturalness can be a useful measure
• Explain why assessment of speaking rate may be important
• Discuss at least three ways in which feelings and attitudes can be assessed ineach of these age groups: preschool, school-age, and adolescents and adults
• Talk about the need for continuing assessment of clients in treatment
HIPAA: Health Insurance Portability and Accountability Act (1996); created national standards to protect the privacy of patient information and still allow access to that information for the safety and proper treatment of patients
Multicultural perspective: An awareness by the clinician of differences in cultures regarding speech, language, and hearing issues as well as differences in styles of interaction between men and women, elders and younger individuals, family and strangers
Bilingual: Having a second language. This applies to individuals who have a single first language that they acquired in infancy and childhood and who then learned a second language or who grew up in a household in which two languages were spoken. Some individuals are multilingual, having learned several languages–either from early childhood or at differing points in time
Empathy: The capacity to understand another’s perspectives, beliefs, and emotions. Having this capacity to some degree allows the clinician to undertake appropriate treatment and to develop trust between herself and the patient–a prerequisite for change
Confronting stuttering: Talking about stuttering, emulating it, and being aware of what’s happening during the moment of stuttering. These activities can be
engaged in by either the clinician or client, when appropriate. These activities and others are thought to reverse the tendency to run away from stuttering that may make stuttering worse
Evidence of reliability: Data that suggest that a procedure or measurement tool produces approximately the same result when used by different individuals or the same individuals at different times
Intrarater reliability: Comparison of results by an individual using a measurement tool at two or more different times
Interrater reliability: Comparison of results by different individuals using a measurement tool
Percentage of syllables stuttered: A common measure of frequency of stuttering obtained by counting the total number of syllables spoken and dividing it into the number of syllables that are stuttered
Types of stutters: The different ways in which an individual may stutter. These include the categories of repetitions, prolongations, and blocks
Duration: The length of time, usually in seconds, that a stutter lasts. From my perspective, this includes the time when forward movement of speech is halted; therefore, the moment of the actual block, prolongation, or repetition is measured as well as the time taken by various starters, postponements, and other secondary behaviors
Severity: Generally a measure of the impediment to communication caused by the stuttering. This may be an overall impression or a compilation of stuttering frequency and duration as well as other behaviors that impede communication
Speech naturalness: The extent to which speech sounds like that of a typical speaker who doesn’t stutter. This measure is useful because sometimes treatment leaves the individual technically “fluent” but sounding overly slow or otherwise odd
Speaking rate: How fast a person talks, usually with short pauses included. (Articulation rate is with the pauses removed.) Speaking rate is most often measured in syllables per minute
Feelings and attitudes: Feelings are the emotions experienced by the person who stutters, especially regarding the experience of stuttering and perceived listener responses. They can vary from one time to the next time. Attitudes are more long-lasting; they reflect the stutterer’s beliefs about how people perceive him and how he perceives himself in regard to his stuttering
Continuing assessment: Measurement and evaluation of changes the client is making/has made over the short and long term
Assessment operates on many levels, like most human endeavors. On one level, there is information gathering, such as interviewing, measuring speech fluency, and administering tests and questionnaires. This requires careful planning, good observation, and thorough analysis. It begins with clients seeking help and often ends with a plan for treatment. On another level, assessment is a personal encounter. It involves getting to know another person and sometimes his family as well, trying to connect to him, and tuning your antennae to pick up the subtle signals he may be sending out about his needs and how you might help him. On this more subjective level, you are becoming aware of the entire person and family, not just the stuttering. Your clients are also getting to know you and sizing up your ability to help them; thus, this first meeting may be the most critical. Although you will want to show an individual client or family that you know about stuttering and understand its treatment, you will want to spend most of your time listening to their concerns and demonstrating your desire to understand them. The two hats you will wear, that of the humanist and that of the scientist, will become a natural part of your wardrobe as you gain more experience.
It is easy to say we must always consider a client’s needs, but it’s hard to put this into practice. One reason is that we can develop expectations that function as blinders. Such expectations affect our perceptions of what our clients want, what caused or precipitated their stuttering, what their priorities are, and many other things. Although I know intellectually that every client is different, I have found a tendency in myself, perhaps increasing as I have become more experienced, to jump to conclusions. I sometimes think, “Ah, yes, I understand this kiddo. So much like that child I saw last month.” You will find this true for yourself too as you work with more and more clients. We must try to listen carefully to what each client says and see each person with fresh eyes.
We must also be cautious about letting referral information, past experience, and biases cloud our ability to see all aspects of the person clearly. We must be wary of simple explanations and quick judgments about which factors are critical for a client. For instance, if parents tell us that they often ask their child to stop and start again when she stutters, that both parents work long hours outside the home, and that dinner is a noisy and confusing time, we should try not to assume that these pressures at home are a major problem for the child. They may be, but other things may be more critical. We need to ask more questions and explore how the child responds in these and other situations before we decide how to begin helping the child and her family.
Sometimes individuals’ or families’ requests differ from what we think they need. An adult may say that she wants “completely fluent speech,” but we know this is not a likely outcome for a person who has been stuttering for 20 or 30 years. Or a family may want us to treat their 3-year-old child without their having to take part in therapy, although our preferred approach for a child this age involves parent participation. I have found it best not to feel I have to resolve such issues during an assessment session. I make no promises but do make a concerted effort to understand what clients and families want and why. My experience has been that after I work with a family or individual for several sessions, we build up enough trust to work together to make the changes that we mutually decide are appropriate.
I remember seeing a young man who came to our clinic from some distance away for intensive therapy. During the evaluation, he made it quite clear that he didn’t want to be treated like a rat in a cage; in other words, he wanted no talk of conditioning, reinforcement, or shaping. In responding to his concerns, I discussed his stuttering with him in terms of what he felt about his stuttering, what he believed his listeners thought, and why he did some of the things he did when he stuttered. Together, we designed an intensive treatment program for him that included plenty of “fluency shaping” and “maintenance” but that made him feel respected as a human and not treated like a laboratory rat. In the process, I also explored with him his concerns about being controlled by others.
In trying to meet a client’s needs, I consider the person as well as the problem. The client, no matter what age, will sense quickly whether a clinician is seeing him as an individual or is only seeing his stuttering. An effective clinician is genuinely interested and empathetic; she accepts failures and backsliding as well as victories and progress. The initial evaluation session is a clinician’s first opportunity to show the client that she accepts him just as he is, without rejection or fear of his stuttering. This atmosphere helps the client begin accepting himself and his stuttering and take the first critical steps toward more fluent speech and effective communication.
All clients should feel they can trust you to protect their privacy and confidentiality. Trust is a vital element of client-clinician relationships. It enables the client to feel that she can safely reveal personal information to you that will help you plan and carry out appropriate treatment. In many cases, the act of expressing feelings in an accepting, secure environment can be therapeutic. For example, a mother whose school-age child was not making progress told her clinician that she was feeling resentment and impatience about her child’s stuttering. She talked at some length about this over several sessions, and the clinician listened empathetically. Once she had released these feelings, her child made remarkable progress. Although this example is more about creating an accepting atmosphere for the child, this parent had to trust the clinician to be accepting of her feelings, as well as not to share this information inappropriately with other family members or the child, or just as inappropriately with fellow student clinicians or others not involved in the child’s care.
Federal and state legislation, such as the Health Insurance Portability and Accountability Act of 1996 (HIPAA), helps clinicians follow guidelines for protecting clients’ privacy. Clinicians should familiarize themselves with these laws and guidelines and ensure that clients give their consent for video recording and observation, and for sharing information about them. You can learn more from the Web site When clients perceive that we are scrupulous in guarding their privacy and confidentiality, we gain a level of trust that enhances therapy. This confidentiality extends to children as well. The bond between a child and clinician will be enhanced if the clinician discusses what information the child is willing to have shared with her parent and what not to share. This is especially relevant for school-age children, who should also be consulted about the extent to which they would be comfortable having their parents involved in treatment.
I should mention that there are a few rare circumstances in which confidentiality may be broken. These are the unusual situations when a client discloses plans to hurt himself or others and instances of child abuse disclosed by the child.
I have been discussing the need to understand and accept everyone who comes to us for treatment as unique. When people who stutter are from other cultures, our task of really understanding them can be more difficult. The 21st century will be a time of more and more migration among cultures and countries. For example, Vermont has recently become home to immigrants and refugees from 23 different countries. States like California, Florida, and Texas have large refugee populations from all over the world. Many refugees have experienced serious trauma that, in some cases, may have precipitated or worsened stuttering. Thus, it is vital for a clinician working with communication disorders to develop a multicultural perspective on assessment and therapy. An underlying principle of this perspective is becoming sensitive to differences in communicative style in other cultures and learning how other cultures view speech and language disorders. You can improve your multicultural sensitivity by reading about cultural issues related to communication disorders in general (Coleman, 2000; Goldstein, 2000; Taylor, 1994) and to stuttering in particular (Conrad, 1996; Cooper & Cooper, 1993; Culatta & Goldberg, 1995; Tellis & Tellis, 2003; Watson & Kayser, 1994). The Stuttering Home Page run by Judy Kuster has a wide variety of interesting papers accessible from a webpage titled “Stuttering in Other Countries/Cultures” (
Tellis and Tellis (2003) make several important points about the importance of cultural knowledge and sensitivity in assessment and treatment. These authors, as well as Watson and Kayser (1994), suggest that finding out about the client’s and/or families’ attitudes about stuttering is crucial. Many of the questionnaires to assess attitudes and feelings that will be presented later in this chapter tap into perceptions of the stuttering behaviors and how much stuttering impedes an individual’s communication, but don’t plumb cultural views of the stigma of stuttering. For example, Tellis and Tellis (2003) report that families from India often feel that stuttering is a reflection on the entire family. This perception may strongly influence family members’ responses to the child’s stuttering, and it may help to discuss the latest information on the etiology of stuttering and the family’s beliefs. The researchers suggest that the clinician ask “open-ended culturally specific questions that address the beliefs, attitudes, and values of the client” (p. 23), in ascertaining how best to work with the client or family.
Some of the multicultural and interpersonal issues relevant to stuttering include:
1. Eye contact. Most treatments for stuttering encourage clients to improve their eyecontact when speaking. A major reason for this is that many people who stutter look away from the listener when they stutter, increasing the perceived abnormality of the symptom and further disrupting communication. This is only true in some cultures, however. In contrast, in some cultures, eye contact with a listener may be inappropriate, depending on the status of the listener and the context. Among some Native Americans, for example, not looking at the listener is a sign of respect. Thus, a person who stutters from such a culture may look away from listeners but not necessarily because of shame or embarrassment. The clinician should become aware of situations in which eye contact while speaking is appropriate and when it is not.
2. Physical contact. During an evaluation or in treatment, many clinicians may touchclients to help them identify points of tension or to signal them to make a change in their stuttering as it is happening. However, many individuals may regard being touched during an evaluation as an invasion of their personal space. It is important to ask permission before touching someone. You might say, for example, “I’d like you to try to catch a stutter and keep it going without finishing the word. Is it OK if I touch your arm to signal you to stay in the stutter?”
3. Nature of reinforcers. Some approaches to treatment use praise as a reinforcer thatis given immediately after a child has spoken fluently. Cultures differ in the amount and type of praise they give children. A clinician I know working in a suburb of Sydney, Australia, a city rich in new immigrants, adapts her treatment contingencies to fit many different cultures. One family from the Middle East was adamantly against giving verbal praise to their child. Instead, they developed a special signal that the father gave to his son to reinforce fluent speech.
4. Family interactions. Children with borderline stuttering are often helped when families change their interaction patterns. One such change that families can make is to speak more slowly and pause between conversational turns when speaking with the child (e.g., Stephanson-Opsal & Bernstein Ratner, 1988). However, in some cultures, particularly in urban areas of the eastern United States, families speak quickly and often overlap each other while talking. For these families, slowing speaking rate and not interrupting each other may seem so unnatural that they are unable to sustain this new interaction pattern. For their children, an operant conditioning approach in daily one-on-one conversations with a parent may be more appropriate.
5. Intentional stuttering. Sometimes I ask the person I am working with in therapy tostutter on purpose, thereby decreasing her tendency to avoid and be afraid of stuttering. But in some cultures, stuttering is regarded so negatively that stuttering on purpose would be unthinkable, at least in the early stages of treatment. It is important for you to become aware of how stuttering is viewed in different cultures and to understand when and where voluntary stuttering might be helpful and acceptable to your clients. Sometimes, the cultural stigma of stuttering makes it difficult for individuals and families to even discuss it.
Our clinic recently treated a young man from China because he wanted to reduce his accent. Only after months of accent reduction treatment was he willing to talk about his greater problem, stuttering. Until we discussed it, he thought he had been successful in disguising it, even though his stuttering was obvious to most of his listeners.
6. Conversational style. Sensitive evaluations and treatment take into considerationnot only the culture’s view of stuttering but also the culture’s style of verbal and nonverbal interaction. Orlando Taylor (1986) described a number of cultural differences in communication style that are relevant to evaluations of stuttering. For example, interruptions of one speaker by another may be expected among African Americans, so trying to change that style of interaction in a family may meet with resistance. In addition, people from African American and Native American cultures may feel uncomfortable responding to the personal questions often asked in an initial interview, and people from a Hispanic culture may feel it is rude to get down to business before greetings and pleasantries are exchanged.
7. Modes of address. The clinician should find out how to address individuals involved in the assessment and treatment, including proper pronunciation. Also, discuss how they’d like to address you. A family from India that I am working with now prefers to address me as “Dr. Barry.”
These cultural considerations are summarized in Table 8.1.
Table 8.1 Cultural and Interpersonal Considerations in Assessment and Treatment

It may not be possible for a clinician to know all relevant aspects of each new client’s culture. But clinicians can be aware of the importance of culture in a person’s response to stuttering, as well as the differences in communication styles between their own cultures and those of their clients. Such awareness can come from reading about a client’s culture and discussing it with the client, if appropriate.
Similar sensitivity should be extended to different social classes within the clinician’s own culture. Understanding and respecting class differences in such areas as vocabulary and values are crucial. Sometimes working with people from other cultures increases our respect for class differences within our own culture. When I worked in Australia, I often attended grand rounds in a Sydney hospital. One particular case presentation involved a working-class Australian woman who had been mutilating herself with needles. Some of the staff and medical residents were highly unsympathetic to her condition, but a psychiatrist, renowned for his work in other cultures, shifted their attitudes. He spoke passionately about how we fail to understand people when we are blinded by our own values and beliefs and that trying to learn about this woman’s circumstances would go a lot further in helping her than our simple condemnations of her self-mutilating behavior.
Some clients will not only be from a different culture or different social class, but they will also speak a different language, one that the clinician may not understand. In this case, an interpreter is necessary. Because interpreters are often from the same culture as that of the client, they may help not only in translating, but also in providing information about important aspects of the culture to aid the clinician’s understanding. In the process of translating sensitive or complex messages, interpreters sometimes need to change the clinician’s message to the client. When a message is rephrased by an interpreter to a more culturally appropriate style, therapeutic interaction will be facilitated. However, if an interpreter doesn’t understand the intent of a question or statement, he may inadvertently convey wrong information. A friend of mine who was working with non-English–speaking Haitian immigrants in Boston understood just enough French to realize that the interpreter was providing wrong information to a client. She rectified the situation by giving the interpreter a brief overview of what she wanted to discuss with the Haitian family and why certain elements were vital, which immediately improved communication.
Special considerations apply when clients are both bicultural and bilingual. Bilingual clients, in fact, are not uncommon; there is evidence of an increased risk for stuttering in bilingual individuals (Howell & Van Borsel, 2011; Karniol, 1992; Mattes & Omark, 1991; Roberts & Shenker, 2007; Van Borsel, Maes, & Foulon, 2001). In these cases, one challenge for clinicians is to determine if the “stuttering” is really stuttering or is simply an increase in disfluency as a result of limited proficiency in a second language. Making this determination may be aided by careful observation of whether there are secondary symptoms (such as eye blinks or signs of increased tension) and cognitive or emotional responses to the suspected stuttering. For example, does the client feel ashamed of her disfluencies? Does she anticipate them? Are they consistently on the same words or same sounds? Another clue is that the disfluencies may be stuttering if there is a history of stuttering in the client’s family.
There is some debate in the literature about the extent to which stuttering occurs in one or more languages of a bilingual speaker. The excellent review of stuttering and bilingualism by Van Borsel, Maes, and Foulon (2001) discusses the evidence on this issue, concluding that although stuttering may occur in one or both languages, it is more likely to occur in both. In some speakers, stuttering may be more severe in one language than another, so that careful analysis of stuttering in both languages will enable the clinician to decide whether to apply treatment to both. Analysis of stuttering in a language not spoken by the clinician is likely to be more accurate if a native speaker of that language, such as a family member or friend of the client, can work with the clinician to identify stutters. In adults, the client herself will be able to help identify stuttering in the language unfamiliar to the clinician. This topic is more thoroughly discussed in the many good chapters on multilingual aspects of stuttering in Howell and Van Borsel (2011).
During an assessment, the clinician has a chance to demonstrate not only her empathy with a client’s feelings but also her mastery of evaluating and treating stuttering. Adolescents and adults who stutter and their family members often come with feelings of frustration, fear, and helplessness. They are looking for someone they can trust and someone who can successfully guide them through the often difficult process of recovery. One of the first things a clinician can do to establish trust and credibility is to show that she not only knows about stuttering, but is comfortable asking questions about it, duplicating it in her own speech, and exploring it empathetically. This provides both clients and family members with an ally, someone who is unafraid of the problem that is so troubling to them.
This process can begin with the clinician asking an older school-age, adolescent, or adult client to teach her how to stutter the way the client himself stutters. This may take some coaching and cajoling, but it will convey the clinician’s willingness to stutter and her interest in the client’s pattern of stuttering. This same credibility can be achieved in the evaluation of a preschool child if the clinician asks about the types of stutters that the child has and demonstrates various possible types such as repetitions, prolongations, and blocks. With a younger school-age child, once she has gotten to know the child a little–this may take a few sessions of working together–the clinician can tell the child she has worked with other kids who stutter but needs to learn about his particular way of stuttering. Then she can ask him if it’s OK if she interrupts him when he stutters to have him show her how to stutter like him. This requires tact, a sense of timing, and even humor to be sure the child feels comfortable confronting stuttering, but it can convey her expertise and thus engender trust.
The clinician’s statements and questions also convey her expertise. For example, as she interviews an older child, she can show that she knows about stuttering by making empathetic comments, such as “Giving reports in front of class can sometimes be hard for kids who stutter.” This allows the child to respond without the pressure of a direct question but also lets the child appreciate that the clinician is someone who has experience with stuttering. When talking with families, the clinician can intersperse questions with such statements as “When children keep repeating a sound that won’t come out, their voices sometimes rise in pitch as the repetition continues.” The family can then confirm whether or not they have noticed this in their child’s speech and at the same time recognize that the clinician is knowledgeable about children’s stuttering. Obviously, these kinds of comments and questions are easier for experienced clinicians, but even beginning clinicians can rely on their reading, their all-too-brief practicum experiences, and their intuition to convey their interest and understanding.
Because it has risks as well as rewards, the approach to interviewing clients and families that was just described should be used carefully. By making comments based on past experience, we may inhibit some individuals and families from telling us about experiences that differ from those offered by the clinician. It is an art to find the balance between showing understanding and leading the witness. As your clinical judgment develops, you will learn which clients will be helped by this approach and when.
I would also caution that demonstrating your expertise should be secondary to acquiring an understanding of clients’ needs. A clinician’s first task is to discern what an individual or family would like from the clinician. The second task is to understand the stuttering problem. In the normal course of accomplishing these two tasks–with attentive listening, empathetic comments, and perceptive questions–the clinician’s expertise will emerge naturally.
Assessment of stuttering behaviors is a broad area that can be divided into several different targets for evaluation, such as frequency, type, duration, and severity. In some situations, it may also be important to assess speech naturalness, speech rate, and concomitant or associated behaviors. The importance of each of these is slightly different depending on the age of the client and the type of treatment you expect to use. Before describing how to assess stuttering, I will clarify what behaviors are considered stuttering. As I mentioned in Chapter 7, a number of authors (e.g., Conture, 2001; Yairi & Ambrose, 1992a) have discussed which types of disfluencies distinguish stuttering from nonstuttering children. Borrowing from their discussions, I have concluded that the following behaviors should be counted as stutters: part-word repetitions, monosyllabic whole-word repetitions, sound prolongations, and blockages of sound or airflow. The latter category (blockages of sound or airflow) can sometimes be quite subtle, occurring in the middle of a word (as in “co-ookie” in which a glottal stop appears to break the word in half) or before a word. I also count successful avoidance behaviors as stutters if they are unequivocally an avoidance. See the section on assessing frequency for a further description of deciding on unequivocal avoidances.
Whenever a procedure is used to assess a behavior or a trait, it is important to know how reliable the procedure is. For example, if a police officer pulls you over for speeding because her radar gun has clocked you going 40 miles per hour in a 25-mph zone, you might want to know how reliable her radar gun was. When this happened to me several years ago, I went to court to contest the ticket. Many factors, I figured, could affect the accuracy of the radar gun’s measurement of my speed: the weather, the age of the gun, and whether it was adjusted properly. Fortunately, the judge asked the officer for evidence of reliability of the radar gun to prove that it could repeatedly, dependably, and consistently measure the speed of a car. Unfortunately, the officer was able to provide the judge with evidence of her radar gun’s recent reliability check, and I shelled out $85 for the fine.
Reliability is obviously an important characteristic of a procedure to measure stuttering. Many factors affect the measurement process, and some of these influences may result in data that are not representative of a client’s true performance. In addition, it appears that stuttering is a particularly changeable behavior, making it difficult to measure. This phenomenon and its consequences are described by Cordes (1994) in her seminal article about reliability:
“Perceptions, judgments, and observations are affected by variables attributed to the observers, to the instrumentation or coding procedures, to the situation or conditions of observation, to the subjects being observed, and to interactions among all of these. Consequently, researchers using direct observation methods are currently expected to provide evidence that their findings are not simply the results of situational influences or observer idiosyncrasies. They are expected, in other words, to provide evidence that their data are reliable” (p. 264).
The same caveat is true for clinical work. Despite our good intentions, observations of clients’ stuttering before treatment and after may be influenced by our desires to see them improve. Measurements may also be affected by random fluctuations in stuttering apart from treatment effects, by the setting in which the client is assessed, by length and type of sample taken, and by the particular dimension of stuttering, such as frequency, severity, duration, or type that is chosen for assessment. It is important for clinicians to learn to assess stuttering reliably and to provide evidence that they have done so. It is also part of the clinician’s responsibilities to know the reliability of the standardized measures he uses to assess stuttering and to choose those measures that are most reliable.
When human judgment is involved, as it always is with measures of stuttering, reliability is checked first by demonstrating that the observer makes the same judgment when observing the same behavior a second time from a video recording, usually several weeks later so that the second observation is fresh and not affected by memories of the earlier judgment. This is called intrarater reliability. Reliability is also checked by comparing the original judgment with the judgment of a second observer who rates the sample independently of the first observer. This is called interrater reliability.
Remeasurement of the data does not have to include the entire sample that is used, although doing so would certainly be the most rigorous approach (e.g., O’Brian, Packman, & Onslow, 2004). It is common for clinical researchers in stuttering to remeasure a randomly selected portion (10–25 percent) of samples taken (e.g., Hakim & Bernstein Ratner, 2004; O’Brian, Packman, & Onslow, 2004). When reliability of judgments is to be established for measurements made on clients who increase their fluency over the course of a treatment regimen, samples should be randomly selected from various points in therapy to include both less fluent and more fluent samples.
Measures of reliability are usually selected according to what behavior is being measured. In situations where evaluation and treatment depend on accurate identification of stuttering moments (such as whether a word or syllable is stuttered or not), reliability can be measured using what is commonly called “point-by-point agreement.” A videotaped sample (e.g., 400 syllables of conversational speech) can be transcribed, and each stutter can be identified and marked on the transcript by an original judge or rater. Sometime later, the rater can return to the sample and again identify stutters by marking a fresh copy of the transcript. The two transcripts are then compared syllable by syllable, and the rater determines how many syllables are agreed upon as stuttered and how many are agreed upon as fluent. This total is termed “number of agreements.” The number of disagreements (syllables that were determined to be stuttered in the first rating but not stuttered in the second rating or vice versa) is totaled and termed “number of disagreements.” The reliability measure is then the number of agreements divided by the total number of agreements plus disagreements, multiplied by 100. Cordes (1994) notes that 80 percent agreement is commonly thought of as the lower limit for a sample to be considered reliable. Figure 8.1 gives an example of a point-by-point assessment of reliability.

Figure 8.1 An example showing how to calculate point-by-point agreement. An initial observer has marked the reading passage by underlining syllables on which stuttering was judged to occur. A second observer has marked the second passage. Point-by-point agreement can be calculated by comparing the total number of agreements with the agreements plus disagreements.
When clinical research is carried out, some authors (e.g., Cordes, 1994) are concerned that point-by-point agreement can be affected by the fact that some agreements might happen by pure chance. To deal with this, it may be useful to report both percentage of agreement for stutters and percentage of agreement for fluent syllables or words. Alternatively, the agreement calculation can be corrected by a procedure that takes into account the effects of chance, such as the kappa statistic (Cohen, 1960; Cordes, 1994).
Point-by-point agreement is appropriate when it is important to judge whether stuttering is present or absent on each syllable. It is also a good tool for new clinicians to use to assess their ability to accurately judge stuttering. However, other procedures are called for when assessment requires quantification rather than presence or absence. An example would be measurement of the duration of stutters. An appropriate measure of reliability would be percent error, obtained by remeasuring at least 10 percent of the data. In this case, it is appropriate to begin by (1) obtaining the absolute differences between each first judgment and each second judgment (change all negative numbers to positive); (2) summing those absolute differences together; (3) dividing by the total number of them to get the average; and finally (4) dividing the average absolute difference by the average of the first judgments. Table 8.2 shows an example.
Table 8.2 An Example of Assessment of Reliability by Calculating Percent
Error of Duration Measurements

Duration of stuttering (in seconds) measured by an observer at Time 1 and remeasured at Time 2. Percent error = 0.2/2.1 = 9.5 percent.
A third method of assessing reliability can be used when measuring the amount of stuttering in cases when point-by-point agreement is not critical. One example would be when you are assessing frequency of stuttering as a measure of week-by-week progress. This procedure involves calculating both the correlation between the first rating and a second rating for multiple samples, as well as test of significant differences between the means of the ratings, such as a paired samples t test. Correlations and t tests should be done for both intrarater and interrater reliability. Correlations should be above 80 percent, and t tests should show no significant difference between the samples. Table 8.3 depicts correlations and t tests for a sample of 12 original and re-rated samples.
Table 8.3 Assessing Interrater Reliability by Calculating Correlations and t Tests

As a final comment about reliability, I would suggest that although different measures of reliability can be used for different purposes, beginning clinicians should establish their reliability using a point-by-point agreement procedure, both during their initial training and to recheck their reliability periodically as they gain more experience. This may help them develop relatively consistent and agreed-upon definitions of what a stutter is and is not.
A summary of reliability measures is given in Table 8.4.
Table 8.4 Measures of Reliability

The size and number of samples depend on the purpose of the assessment. In a first assessment, it would be wise to have at least two samples: one recorded in the clinic and one recorded in the client’s typical environment. Before I see a preschool child for an evaluation, I ask parents to send in a video of the child in conversation at home. Video recording is common in some homes so that the presence of a video camera will probably not make most children self-conscious. I sometimes ask parents to leave the camera on a tripod in a familiar place for several days so the child is used to it when the recording is actually done. When video recording is not possible, audiotaping is still useful. With a school-age child, a sample collected in the school would be important; practically speaking, a sample could be most easily recorded in the therapy room. A second sample from home would also be very helpful, but it is not always obtainable. When evaluating an adolescent or adult, I recommend that a sample be taken in the treatment room and a sample be taken from work or home. It is often convenient for adolescents or adults to audiotape telephone conversations. I often urge clients to buy a small digital audio recorder for recording work done outside therapy; Sony has one for $20.
An important consideration in obtaining samples is that stuttering varies. It differs in frequency and severity from month to month, week to week, day to day, and from situation to situation within the same day. Such variability affects both children and adults but is most apparent with younger stutterers. Sometimes a preschool child is stuttering severely, and then three weeks later during the evaluation, the child is entirely fluent. Therefore, it is important to discuss with the client or family whether the sample you have obtained is representative of the stuttering and if not, whether more samples should be taken, maybe in other situations and at other times.
After the initial sample, when further assessment is done to measure progress in therapy, it is crucial to ensure that any reduction in stuttering is not confined to the therapy room. Thus, ongoing assessments should include measures taken in the client’s real world, outside the treatment situation.
For any sample in which severity of stuttering is to be rated or any sample for research purposes, videotaping is essential. Many subtleties of stuttering would be missed if only an audiotape were used; thus video recording allows better assessment of observer reliability than audio recording. Sometimes online (while the client is talking) scoring can be done without either video or audio recording. For example, online scoring is appropriate when the clinician samples frequency of stuttering at the beginning of every session for clinical rather than research purposes.
The length of the sample must be long enough for it to be representative of the speaker’s typical stuttering. A sample that’s too short won’t include enough stuttering to see the range of severity and types of stuttering, and a sample that’s too long would take time away from other assessment activities and would be tedious to score. For a client who reads, I usually like to take a sample of 300 to 400 syllables of conversational speech (where there is likely to be more variability) and 200 syllables of a reading passage (where there is likely to be less variability). Using a typical figure of 1.5 syllables per word (Williams, Darley, & Spriestersbach, 1978), these samples would be equivalent to approximately 200 to 265 words and 130 words, respectively.
When obtaining a reading sample, it is important to ensure that the reading passage is at or below the client’s reading level. A client’s stuttering is likely to worsen when reading a passage that is difficult, giving a false impression of typical stuttering during reading. Reading passages in the Stuttering Severity Instrument-4 (SSI-4) (Riley, 2009) are designed for third, fifth, and seventh grade levels, as well as for an adult reading level. You can also write your own passages and check them for grade level using the Tools option on Microsoft Word, which uses the Klesch-Kincaid Reading Level statistics, or you can use the Fry Readability Graph available at
When obtaining a speaking sample, it would be wise to select topics that are not emotional unless it is desirable to elicit a maximal amount of stuttering, as you might do with a client who says she stutters but is not demonstrating any during the evaluation. I usually ask children and adolescents to talk about their favorite weekend or afterschool activities, sports, hobbies, or pets. With adults, I ask them to talk about their favorite activities, sports, hobbies, work, or school.
When making a formal assessment or when first learning to assess stuttering, it is very useful to make a written transcript of the spoken material, including all words and even those nonmeaningful utterances, such as “uh.” However, you should not indicate on the transcript which syllables are stuttered, so that you, at a later date, or another rater, on a separate occasion, can rescore a copy of the transcript to check for reliability without being influenced by the notations indicating which syllables were stuttered. Using your recording of the spoken material and an unmarked transcript, you can note where the stutters are, with details of how the individual stuttered. Write out each element of a repeated sound or syllable, the sounds that were prolonged, and the sounds on which blocks occurred. Describe escape and avoidance behaviors accompanying each moment of stuttering. Mark those moments of stuttering that seem longer than most. For a complete assessment, you will want to return to the longer stutters and time how long each one was to determine the average length of the three longest stutters. You will also want to count the words or syllables spoken, although it is often most accurate to count syllables from recordings because some speakers omit syllables in longer words. This can be done with software, such as the Computerized Scoring of Stuttering Severity (CSSS) software that accompanies the SSI-4.
Assessing Frequency
Frequency of stuttering is a simple, reliable measure (Andrews & Ingham, 1971) that can be used for a variety of purposes. It is important in an initial assessment to help distinguish a normally disfluent child from a child with borderline stuttering. It is a vital part of composite ratings, such as the SSI-4 (Riley, 2009), that provide a multidimensional view of stuttering. Frequency of stuttering is also useful as a “snapshot” measure of progress during treatment. In the first place, it is highly correlated with severity (Young, 1961). If used alone, however, frequency has the limitation that it doesn’t reflect the duration of stutters or physical tension associated with stuttering. Decreases in these variables are often signs of improvement.
Frequency of stuttering is most commonly reported as percentage of syllables stuttered, although some use percentage of words stuttered or number of stutters per 100 words. I prefer to use percentage of syllables stuttered, following the logic of Minifie and Cooker (1964), because it can capture instances when a speaker stutters on more than one syllable of a multisyllable word. Moreover, when counting syllables and stutters online, syllables can be counted more easily than words by counting the syllable beats as the client talks. When counting stutters, I assume that each syllable can be stuttered only once. Thus, multiple repetitions, like “Where is my ba-ba-babasketball?” are counted as only one stutter. “Where is my…my…uh…well ba-ba-baba-basketball?” is also one stutter, because I assume that the repetition of “my” and interjections of “uh” and “well” are postponements associated with the stutter that was anticipated and actually occurred on “basketball.” If a speaker appears to have a habit of using a particular word or sound as an avoidance behavior, I will count a word as stuttered even if no overt stuttering occurred. For example, a speaker may say “Where is my…uh…uh…uhbasketball?” In this case, the speaker appears to be using “uh” to postpone starting the word “basketball” on which he anticipates stuttering. And he keeps saying “uh” until he feels he can say “basketball” fluently, then rushes to say “basketball” after saying the last “uh.” When I am fairly certain that a speaker has used a sound or word as a (successful) avoidance behavior like this, I count the word as stuttered. Note that I do not count each utterance of the sound or word that is used as an avoidance; instead I count the next word in the utterance as stuttered. When I am in doubt about whether an avoidance has occurred on a word, I count it as fluent.
When assessing the speech of someone who can read, I find it helpful to compare the frequency of stuttering in reading to that in speaking. If stuttering is markedly greater in the reading task, this may be because the speaker is avoiding words he expects to stutter on in the speaking task, but he can’t do this when reading. In most cases, I talk about my hypothesis with the client to see if he agrees.
A variety of instruments designed for counting stutters are available. A free online counter that counts stuttered syllables, fluent syllables, and calculates percent syllables stuttered and speech rate is available at (recommended by Julie Pera). You could also use two electronic hand-held counters–one in each hand, available at for under $15 (recommended by Jennifer Code). Some clinicians using the Lidcombe Program for Early Intervention (Chapter 10) prefer to use a hand-held counter called a TrueTalk ( My own preference is to use a TrueTalk, which counts syllables and stutters, as well as time during which the counter buttons are pressed. Output includes total syllables spoken, percent syllables stuttered, time elapsed, and stutters per minute. This device can be used with one hand and held discretely at your side when assessing a client’s speech while you are standing, sitting, or walking.
Assessing Types of Stutters
When assessing the speech of preschool children, it is often useful to count the total number of disfluencies, both those that are considered types of stutters and those considered normal. As you will remember from Chapter 7, disfluencies that are not considered stutter-like include multisyllable word repetitions, phrase repetitions, interjections, and revisions in which a phrase is incomplete. When both types of disfluencies (stutter-like and not) are counted, you can use the proportion of total disfluencies that are stutter-like to help you decide whether a child is stuttering or normally disfluent. As I indicated in Chapter 7, Yairi (1997a, 1997b) surveyed a number of studies and proposed that if less than 50 percent of a child’s disfluencies are stutter-like, the child is more likely to be normally disfluent. Caution must be used with any single measure used alone. Conture (2001) noted that a child he had recently evaluated was, in his opinion, stuttering severely even though the child’s proportion of stutter-like disfluencies was only 34 percent of the total disfluencies. Clearly, Conture had relied on several other measures of stuttering in concluding that the child was a severe stutterer.
Another measure involving the type of disfluencies that a child produces is the number of stutter-like disfluencies per 100 words. In summarizing his findings on disfluencies in stuttering and nonstuttering children, Yairi (1997b) noted that children who stutter have more than three stutter-like disfluencies per 100 words, whereas normally disfluent children have fewer. In this same chapter, Yairi reviewed research about the gradual decline in some types of disfluencies as children grow older. Perhaps the most important finding is that part-word repetitions show a steady decline in normally disfluent children by age 4 and thereafter. Thus, if a child shows a plateau or increase in part-word repetitions in later preschool years, the child may be showing stuttering rather than normal disfluency.
Assessing Duration
In a thorough assessment, measures of the duration of a client’s longest blocks can give us important information about how much stuttering may be interfering with communication. Van Riper (1982, p. 208) noted in his inimitable prose that “The duration of the individual moments of stuttering is one of the basic components of any adequate index of severity. Like tapeworms, longer stutterings are worse than shorter ones.”
A common practice is to average the duration of the three longest stutters in a speech sample (Myers, 1978; Preus, 1981; Riley, 2009; Van Riper, 1982). One way to do it is use a digital stopwatch while watching a videotape of the client speaking. With a little practice, you can turn the stopwatch on at the moment the stutter begins and turn it off when it ends and measure the moment of stuttering to the nearest halfsecond. Any delays in starting the stopwatch at the beginning of stutters are compensated for by similar delays when you stop it at the end. I recommend using duration as part of a more complete assessment of severity, such as the SSI-4 (Riley, 2009), when making an initial assessment of a client’s progress and when you want to give a detailed description of a client’s stuttering in a report. The software accompanying the SSI-4 provides a means to automatically calculate the mean of the three longest stutters by holding down the mouse key for the duration of each stutter as it is being counted.
Assessing Secondary Behaviors
Stuttering feels like being in the grip of an unseen hand damming up the flow of your speech. Or as one of my young clients said, it is like having “a rock jammed in your throat.” You struggle to keep going, squeezing your lips, blinking your eyes, or twisting your shoulders in the process. Such behaviors add to the abnormality of stuttering and reflect an important aspect of its development. Reducing or eliminating these behaviors may be a vital goal for therapy.
Secondary behaviors are also referred to as “concomitant,” “associated,” or “accessory” behaviors. They are most often escape behaviors that are used to break out of a stutter once it has started, but secondary behaviors may also be avoidance behaviors that are used in an attempt to keep from stuttering (see Chapters 1, 4, and 7 for further discussion of these terms). These behaviors may be physical movements (e.g., eye blink), extra sounds (e.g., “uh”), or changes in the way speech is produced (e.g., pitch rise). They are often signs that stuttering has progressed to a more advanced stage (i.e., escape behaviors distinguish beginning from borderline stuttering), but they may in a few cases appear very close to the onset of stuttering.
Conture (2001) briefly reviewed the limited research on secondary behaviors and noted that most are just more frequent and more exaggerated versions of behaviors seen in normal speakers. Zebrowski and Kelly (2002) suggested that the most common behaviors involve the eyes, particularly blinking, squeezing, lateral and vertical eye movements, and loss of eye contact. These authors and Shapiro (1999) also pointed out that the presence of secondary behaviors can be an important diagnostic sign that may distinguish normally disfluent children from those who are beginning to stutter.
Some clinicians enumerate these secondary behaviors as part of their assessment, particularly when they will use a treatment approach that helps the client gradually modify her stuttering behaviors. Standardized measures, such as the SSI-4 (Riley, 2009), include ratings of these behaviors as part of an overall severity assessment. We will consider this assessment next.
Assessing Severity
Measures of severity may be the most clinically relevant assessment of overt stuttering behaviors. Severity reflects an overall impression that listeners may have when they listen to an individual who stutters. Thus, it is an important measure for assessing the outcome of treatment. It is also an important yardstick of progress during therapy because many treatments gradually reduce the severity of stuttering rather than eliminate it.
The most commonly used measure of severity is the SSI-4, which was first published in the Journal of Speech and Hearing Disorders (Riley, 1972). A recent version, the SSI-3 is illustrated in Figure 8.2, and the most current version is available with forms and a manual from PRO-ED ( In my mind, it is the best measure of severity available, but like its predecessors, the SSI-4 has some drawbacks. The sample of children and adults on which it was normed is not well described, its reliability is not strong, and its validity has not been convincingly demonstrated (McCauley, 1996). Despite these limitations, the SSI is easy to use and captures the severity of overt stuttering behaviors as a composite of three important dimensions: frequency, duration, and physical concomitants. The SSI-4 is one of the few measures of stuttering that has standardized procedures for gathering and scoring speech samples and is the only measure that includes the three dimensions just cited.

Figure 8.2 The Stuttering Severity Instrument-3.
The total overall score for the SSI is the sum of the three subcomponents measured.
1. Frequency is assessed as the percentages of syllables stuttered on a speaking taskand a reading task. For nonreaders, the speaking task is given twice the weight in the scoring procedures. Riley originally used percentage of words stuttered but currently uses the percentage of syllables stuttered, which is converted to a “task score” on the form.
2. Duration is assessed by measuring the length of the three longest stutters, calculating their mean duration, and finding the appropriate “scale score” on the form.
3. Physical concomitants are assessed by adding the scale values of each subcomponent (i.e., distracting sounds, facial grimaces, head movements, and movements of the extremities) and deriving a total score.
The values for frequency, duration, and physical concomitants are then added together to provide a total overall score. Percentiles and severity ratings (e.g., mild, moderate, and severe) based on total overall scores are given on the form. Clinicians should carefully read Riley’s directions on administering this measure in the manual of the SSI-4 before using it.
Clients should be video recorded, and the SSI should be calculated from the recording because duration measures and assessment of physical concomitants cannot be done easily online, and the frequency count will be more accurate if equivocal stutters are replayed repeatedly until a decision can be reached. The CSSS software accompanying the SSI-4 allows computer-aided calculation of stuttering frequency and duration for the overall severity score.
A relatively new measure of severity, the Test of Childhood Stuttering (TOCS) (Gillam, Logan, & Pearson, 2009), can be used with children ages 4 to 12. It can be obtained from PRO-ED ( The TOCS consists of several subparts:
1. A Speech Fluency Measure which is comprised of (a) rapid picture naming; (b) modeled sentences; (c) structured conversation; and (d) narration.
2. An Observational Rating Scale to be used by the clinician, teacher, or caregiver.This component provides information from the observer about (a) how often the child has various stuttering behaviors, and (b) how often the child has negative responses to his stuttering, such as showing such secondary reactions as concomitant physical behaviors and avoidance of speaking.
3. A Supplemental Clinical Assessment which allows a more detailed analysis of thestuttering frequency, duration, types, and associated behaviors, as well as speech naturalness. This measure can help to decide if the child stutters and how severe the child is. It can also be used for pre- and post-treatment assessment.
In a review of TOCS in Mental Measurements Yearbook, Shapely and Guyette
(2010) comment favorably on the instrument’s validity and reliability, but suggest caution in interpreting the test’s index scores and percentile ranks because of limited sample sizes used in standardization and in validity and reliability assessment.
Another measure of severity, which captures frequency, duration, and perhaps secondary behaviors, is the Scale for Rating Severity of Stuttering (Johnson, Darley, & Spriestersbach, 1952, 1963; Williams, 1978). This early scale, shown in Figure 8.3, is more subjective than the SSI, relying on an overall impression of a speech sample to rate the sample with one of eight values (0–7). Raters are encouraged to treat each of the eight intervals between the scale values as equal, although there is some debate about whether this is truly an equal-interval scale (Berry & Silverman, 1972). Although it has been shown to be reliable when a group of raters is used, the reliability of the Scale for Rating Severity of Stuttering for use with single raters is questionable. Williams (1978) cautions that the scale gives only a rough measure of severity because of its limitations. But he also notes that it has clinical utility because it captures the listener’s impression of a client’s speech and may therefore convey information about what the client faces every day when he is speaking. Ratings by a number of real listeners in the client’s environment, as well as the client herself, would increase the value of this information. Any scale, including this one, has real risks if used as a single measure of therapeutic progress by a clinician. Unconscious bias and familiarity with the client may lead to improved ratings in the absence of change. Progress should be assessed with a variety of tools, including the SSI-4.

Figure 8.3. Scale for rating severity of stuttering.
A fourth measure of severity is the Lidcombe Program’s Severity Rating Scale, which was developed by Onslow, Costa, and Rue (1990) as part of an operant treatment program for preschool children. This is simply a 1-to-10 scale that parents use to make daily ratings of their child’s stuttering (1 = no stuttering, 2 = extremely mild stuttering, through 10 = extremely severe stuttering). The scale, in a format that allows for a week’s ratings, is shown in Figure 8.4. At the beginning of treatment, parents are trained to accurately rate their child’s severity using observations of the child’s speech in the clinic. The clinician and parent compare their ratings and discuss any differences between them until the parent’s ratings are within one scale value of the clinician’s. Throughout treatment, a sample of the child’s speech that is long enough to ensure that any stuttering in the child’s speech is observed is taken at the beginning of each clinic meeting. Both the clinician and the parent rate this sample, ensuring continued agreement.

Figure 8.4. The Severity Rating Scale. Rate the speaker on a 10-point scale, where 1 = no stuttering and 10 = extremely severe stuttering (the worst stuttering the speaker has produced) for the entire day. Put an X in the appropriate box at the end of each day.
This severity rating scale has also been used with school-age children who stutter. These older children often rate themselves in a version of the Lidcombe Program developed for older children. Research on this severity rating scale has shown it to be a valid and reliable tool for conveniently obtaining information on a child’s stuttering outside of the treatment environment (Onslow, Andrews, & Costa, 1990; Onslow, Harrison, Jones, & Packman, 2002).
In recent years, clinical scientists have been concerned that treatments that produce fluency may not always result in natural-sounding speech. As Schiavetti and Metz (1997) warned, “Some stutterers may reduce their number of stutters at the expense of a speech pattern that is stutter-free but not really fluent.” Thus, some stuttering treatments may get rid of stuttering but leave an individual with speech that sounds odd, unusual, or unnatural. Martin, Haroldson, and Triden (1984), one of the first investigative teams to report on this problem, found that unsophisticated listeners rated the stutter-free speech of individuals who stutter speaking under DAF (delayed auditory feedback) as significantly more unnatural than the general speech of nonstutterers. Ingham, Gow, and Costello (1985) used the same rating scale and found that the fluent speech of treated stutterers was judged to be more unnatural than that of nonstutterers. Both investigations used a nine-point, equal-appearing interval scale to rate speakers based on judges’ intuitive sense of what sounded “natural,” speech naturalness. The judges in these and most subsequent studies exhibited satisfactory levels of interrater reliability and agreement, although individual rater reliability was only marginally satisfactory.
Clinically, we need to be sure that clients sound as natural as possible after treatment. Otherwise, they are likely to abandon their fluency skills in favor of old, familiar stuttering patterns because of their own and listeners’ negative reactions to their post-treatment speech. Can we rate our clients’ naturalness reliably? Schiavetti and Metz (1997) indicated that clinicians who have learned to be consistent raters of speech naturalness may rely on the relative values of their ratings. Thus, they can judge when a client sounds less natural than other clients they have treated and take appropriate steps to improve that client’s naturalness before releasing her from treatment. The SSI-4 incorporates a naturalness rating as part of the assessment.
Many clinicians believe that speaking rate often reflects the severity of stuttering (e.g., Shapiro, 1999; Starkweather, 1985, 1987). Van Riper (1982) described studies that found correlations that ranged from 0.68 to 0.88 between reading rate and severity. If a client’s speaking rate is well below average for her age, communication will be affected; listeners may become impatient or lose the thread of what the speaker is saying. Speech rates that are too fast will also affect communication. A subgroup of individuals who stutter also have the disorder of cluttering, which is rapid, often unintelligible speech (see Chapter 15). Thus, it is useful to measure the client’s rate in standard speaking and reading tasks. Table 8.5 gives average speaking rates for children and adults.
Table 8.5 Speaking Rates for Children and Adults

clinician’s preference. Some clinicians find it easier to calculate rate by using words per minute because words are easily observable units on the page. Others note that syllables per minute can be calculated more rapidly than words because clinicians can use the “beat” of syllables to count them online (i.e., while a speaker is talking). The syllables-per-minute approach also accounts for the fact that some speakers use more multisyllabic words than others and might be penalized because such words take longer to produce than one-syllable words. I recommend using syllables for these reasons.
No matter which method is used, the following rules can be used for counting words or syllables. Count only those words or syllables that would have been said if the person had not stuttered. Thus, if a person says, “My-my-my, uh, well my name is Peter,” this should be counted as four words or five syllables because it is apparent that the extra instances of “my,” the “uh,” and the “well” are part of the stuttering. If a person says, “When I went to Boston, I mean when I went to New York…,” and it does not appear that the person was postponing or using a “trick” to avoid stuttering, this would be counted as 13 words or 14 syllables because stuttering did not interfere with the utterance. Only true words (or syllables in true words) are counted; “uh” or “um” are not counted. “Oh” or “well” are counted, unless they are used as a postponement, starter, or other component of stuttering. These distinctions may seem difficult to remember, but the main rule of thumb you should use is that you are counting syllables or words that convey information to the listener.
When syllables per minute are calculated, it is often easiest to use an inexpensive calculator to count syllables cumulatively as they are spoken, although this takes some practice. Before the speaker begins, push the “1” key, then the “+.” When the speaker starts speaking, press the “=” key for each syllable spoken or read, and the cumulative total will appear in the readout window. It is easier to count syllables by reading a transcript of the conversational sample aloud slowly and pushing the “=” key for each syllable spoken; inexperienced raters should learn to count syllables first from a transcript. Experienced raters can assess conversational speech rate directly from recordings by pressing the “=” key for each syllable spoken. Some calculators will count cumulatively when the “1” is pressed, followed by repeated presses of the “+” button; I have not found an expensive calculator that will count cumulatively (a cheap one will), but the calculator that came with my PC software will count cumulatively with the 1, +, = maneuver.
Some clinicians have found that they are able, with practice, to count syllables per minute as the client is speaking by using graph paper with small boxes. As the client is talking, they put a dot in each box for each syllable spoken. They also use this method to assess frequency of stuttering, by putting a check instead of a dot for each syllable stuttered.
When words per minute are calculated, a transcript is made of a client’s five-minute sample of conversational speech, and her five-minute reading sample is marked to indicate where she finished. The total number of words is counted, and this figure is divided by five to give a per-minute conversation or reading rate.
It is important to measure these samples accurately with a stopwatch or another timing device. In measuring the amount of speaking time in a conversational sample, I stop the stopwatch whenever the client is not talking but allow it to run during moments of stuttering. Short pauses of less than two seconds are incorporated into the five minutes, but formulation pauses longer than two seconds are excluded. With a little practice, starting and stopping a stopwatch during pauses and turn-switching becomes easy and natural.
The CSSS software that comes with the SSI-4 allows the clinician to count number of stutters and number of fluent syllables and assess total sample duration in seconds. By totaling the stuttered and fluent syllables and dividing that by the sample duration in minutes (minutes + number of seconds divided by 60) you can get overall speech rate in syllables per minute.
The feelings or emotions of individuals who stutter, as well as their beliefs and attitudes about themselves, about communication, and about stuttering, are all components of stuttering. For most people who stutter, the experience of stuttering and the reactions of others to their stuttering have a notable effect on their behavior and on their response to therapy. Therefore, assessment of these aspects of stuttering is important. In this section, I focus on formal measures of feelings and attitudes that can be administered throughout therapy to assess progress. As clinicians gain more experience, they will develop informal procedures to supplement formal measures. I will describe these in the next chapter, as well as in appropriate treatment chapters.
Assessment of Preschool Children
There are not many instruments to assess feelings and attitudes of preschool children who stutter, but two are worth noting. The first is the KiddyCAT (Hernandez, 2001; Vanryckeghem, Brutten, & Hernandez, 2005; Vanryckeghem, Hernandez, & Brutten, 2001), which was designed to assess communication attitudes of preschool children who stutter. It consists of 12 yes/no questions asked by the clinician after some practice items. An example is “Do you think that Mom and Dad like the way you talk?” The questions are asked in a play environment, such as putting a marble in an egg carton after each response. Higher scores indicate more negative attitudes about speech. In a sample of 45 children who stutter (ages 0–3 to 5–6) and 63 children who don’t (ages 2–3 to 3–6), the mean scores were 4.35 (SD = 2.78) for children who stutter and 1.79 (SD = 1.78) for children who don’t. This difference is statistically significant (p <.001).
Another assessment tool for attitudes and feelings of preschool children who stutter takes an indirect approach by surveying parents about the impact of stuttering on the child and themselves (Langevin, Packman, & Onslow, 2010). The survey–The Impact of Stuttering on Preschoolers and Parents (ISPP)–consists of 20 questions covering (a) child-related questions (e.g., “Has your child ever been frustrated when stuttering?”); (b) questions about playmates (e.g., “Has your child ever been teased by other children about his stuttering?”); and (c) parent-related questions (e.g., “Has your child’s stuttering ever affected you emotionally?”). One of the purposes of this tool, according to the authors, is to help decide whether to enroll the child in treatment or to delay treatment to see if stuttering resolves. This survey awaits further testing to assess validity and reliability. The ISPP is shown in Figure 8.5.

Figure 8.5. The impact of stuttering on preschoolers and parents.
A number of clinicians have suggested that a preschool child’s sensitivity or reactivity to new situations may be an important consideration in therapy and possibly predictive of chronicity (Conture, 2001; Guitar, 1998). I have found that the Behavioral Style Questionnaire (BSQ) (McDevitt & Carey, 1978, 1995), which is administered to parents, provides some information on this dimension. Some research indicates that the BSQ may be able to identify those children who have a more inhibited, sensitive temperament (Anderson, Pellowski, Conture, & Kelly, 2003; Conture, 2001). The BSQ has relatively high mean test-retest reliability for the 3 to 7 years scales (0.81) and a moderate internal consistency for this age range (0.70) (McDevitt & Carey, 1995).
Assessment of School-Age Children
There are several tools available for assessing attitudes about speaking or about stuttering in children. Figure 8.6 depicts a scale that a graduate student and I developed many years ago, called the A-19 Scale (Guitar & Grims, 1977). It consists of questions that we have found will distinguish between children who stutter and children who do not. Thus, if treatment is effective, a child’s attitude about communication may change, although this has not been established by research.

Figure 8.6. A-19 Scale of Children’s Attitudes by Susan Andre and Barry Guitar (University of Vermont; reprinted from Susan Andre, with permission).
In addition to the A-19, the Communication Attitude Test (CAT), which was developed by Brutten and his colleagues, has been shown to be a reliable and appropriate tool to measure the attitudes of children and adults who stutter. In a study by De Nil and Brutten (1991) 70 children who stuttered and 271 children who didn’t (ages 7–14) were administered the CAT. The children who stuttered had significantly more negative communication attitudes (p <.01), and this difference increased with age. Vanryckeghem and Brutten (1997) replicated this finding with children (ages 6– 13). In 1992, Vanryckeghem and Brutten affirmed the CAT’s test-retest reliability using 44 school-age children; results indicated correlations of 0.86, 0.81, and 0.76 for retesting after 1, 11, and 12 weeks, respectively, indicating that it is appropriate to administer it to a child several times. In a study of 143 children who stuttered (ages 7– 13), Vanryckeghem, Hylebos, Brutten, and Peleman (2001) examined the relationship between negative attitudes as measured by the CAT and negative emotions elicited by the questions on the CAT, as measured by a 1-to-5 scale filled out by the children. The authors found a high positive correlation (r = 0.89) between negative attitudes and negative emotions. Both negative attitudes and negative emotions increased with age. In summary, the CAT is a well-researched tool that can be used to determine the presence of negative attitudes in individuals ages 6 and older. It is now part of a larger assessment battery (Behavior Assessment Battery for School-Age Children Who Stutter, Brutten & Vanryckeghem, 2007), which also contains the Speech Situation Checklist that evaluates a child’s reactions to a range of situations, as well as the Behavioral Checklist which assesses a child’s coping responses to his disfluency.
It is important that a trusting relationship with the child has been developed before administering either the A-19 or the CAT to a school-age child.
The Overall Assessment of the Speaker’s Experience of Stuttering (OASES) is a questionnaire designed to assess the impact of stuttering on a person’s life. It is based on the World Health Organization’s International Classification of Functioning, Disability, and Health (World Health Organization, 2001) and focuses on feelings about stuttering, reactions to stuttering, communication in daily situations, and the extent to which stuttering interferes with daily living. A new version is adapted for use with children ages 7 to 12. A more detailed description is given in the next section on adolescents and adults, and a sample for adults is shown in Figure 8.7.

Figure 8.7. Overall assessment of the speaker’s experience of stuttering. (Reprinted with permission from NCS Pearson, Inc. © 2010.)
Although not strictly a measure of attitude, the Teachers Assessment of Student Communicative Competence (TASCC) (Smith, McCauley, & Guitar, 2000), which is depicted in Figure 8.8, is useful in assessing a child’s communicative functioning in the classroom. I have listed it here because one of the subscales purports to measure approach/avoidance in the classroom based on questions about the child’s class participation and volunteering to talk. Other areas that the teacher rates the child on include intelligibility, comprehension, appropriateness of communication, and pragmatic/nonverbal communication skills. I have found the TASCC to be helpful in getting information about how a child’s communication is changing over the course of treatment. The measure was tested on 69 students in grades 1 through 5 in Maine,
Vermont, Texas, Virginia, and Idaho and showed high internal consistency. Cronbach’s coefficient alphas (a measure of a test’s reliability) for the five subscales ranged from 0.77 to 0.95, respectively, suggesting that the TASCC items were related measuring a similar construct (Smith, McCauley, & Guitar, 2000). Redundancy analysis was used to remove redundant items and combine some that were similar, leaving a 50-item scale.


Figure 8.8. Teacher assessment of student communicative competence.
Further pilot work (Sequin, 1999) on the TASCC was conducted with 14 children who stuttered, paired by gender and ethnic background with 14 who did not stutter.
Scoring of each pair of participants was obtained from teachers of children in grades 1 to 5; the participants included eight Caucasian pairs, two Hispanic pairs, and four African American pairs. Three of the pairs were females and 11 were males. The children who stuttered had significantly lower communicative competence scores (p =.0001) than the control children; the “approach/avoidance attitude” subscale showed the greatest difference between the groups. A second pilot study (Pierson, 2004) compared TASCC ratings of eight children who stuttered and eight children who did not (grades 1–5) matched for age, gender, grade, cultural background, and academic performance. Again, the children who stuttered had significantly lower communicative competence scores (p =.002). The approach/avoidance attitude subscale was not found to be significantly different between the groups, but these three scales were, with the differences between the groups in this rank order: intelligibility > appropriateness of communication > clarification/repair (of output)/comprehension (of input) > pragmatic/nonverbal communication. Clinically, I have found it useful to know what subscales are most deviant for a child. In one instance, a child who was rated deviant on the intelligibility subscale benefited from learning to speak more slowly and loudly, even when he stuttered.
Assessment of Adolescents and Adults
A variety of questionnaires can be used to assess various aspects of a stutterer’s feelings and attitudes about communication and stuttering. I typically use the Modified Erickson Scale of Communication Attitudes (S-24) (Andrews & Cutler, 1974) to obtain information about a client’s communication attitudes (Fig. 8.9). This questionnaire has been normed on both stutterers and nonstutterers. A colleague and I (Guitar & Bass, 1978) studied a sample of 20 individuals treated by a fluency-shaping program and found that if communication attitude as measured by the S-24 does not change during treatment, the likelihood of relapse within 12 to 18 months increases. Ingham (1979) disputed this finding, but Young (1981) confirmed it using a reanalysis of the original data. Later data by Andrews and Craig (1988) also supported the relationship between normalizing attitudes on the S-24 and long-term treatment outcome.

Figure 8.9. Erickson S-24 Scale of Communication Attitudes. (Reprinted from Andrews, G. & Cutler, J.
(1974). Stuttering therapy: The relation between changes in symptom level and attitudes. Journal of Speech and
Hearing Disorders, 39, 312–319, with permission. Copyright 1974, American Speech-Language-Hearing Association.)
I also use a questionnaire to assess a client’s tendency to avoid stuttering, which is the avoidance scale of the Stutterer’s Self-Rating of Reactions to Speech Situations (SSRSS) (Johnson et al., 1952); this questionnaire assesses a client’s tendency to avoid specific speaking situations (Fig. 8.10). Research suggests that clients with avoidance scale scores higher than 2.56 before treatment may be more likely to have appreciable levels of stuttering one year after treatment with fluency-shaping therapy than clients with lower scores (Guitar, 1976). Thus, I suggest that clinicians use a client’s avoidance scale score to guide them in choosing whether to focus more on ways of enhancing fluent speech or to combine shaping fluency with an approach that modifies stuttering as well as the fears and avoidances associated with stuttering.

Figure 8.10. Stutterer’s Self-Rating of Reactions to Speech Situations. (Reprinted from Johnson, W., Darley, F., & Spriestersbach, D.C. (1952). Diagnostic manual in speech correction. New York: Harper & Row, with permission. Copyright 1952 by Harper & Row. Copyright renewed 1980 by Edna B. Johnson, Frederick L.
Darley, and Duane C. Spriestersbach.)
Sometimes I also use the Perceptions of Stuttering Inventory (PSI) (Woolf, 1967) to examine a stutterer’s perception of the presence of struggle, avoidance, and expectancy of stuttering (Fig. 8.11). Woolf suggests that the PSI can be used to help a stutterer view her problem more objectively, to develop treatment goals, and to assess progress. I find that the avoidance section of the PSI complements the avoidance scale of the SRSS because the SRSS focuses more on situations, whereas the PSI deals more with stuttering behaviors.

Figure 8.11. Perceptions of Stuttering Inventory. (Reprinted from Woolf, G. (1967). The assessment of stuttering as struggle, avoidance, and expectancy, British Journal of Disorders of Communication, 2, 158–171, with permission.)
Another measure of attitude that has been shown to predict long-term outcome is the Locus of Control of Behavior Scale (Craig, Franklin, & Andrews, 1984), which assesses the extent to which a client believes he controls his own behavior (i.e., whether the control is “internal” or “external”) (Fig. 8.12). Scoring adds the points for each item, and higher scores reflect greater perceived “externality” of control. Because the values of items 1, 5, 7, 8, 13, 15, and 16 are reversed to minimize the effect of social desirability in responding, the scores on these items are transposed (e.g., you change a 5 to a 0, a 4 to a 1, and so on) before totaling the score. This scale is given just before treatment and again immediately after treatment. Studies have shown that clients who did not decrease their locus of control scores more than 5 percent from pretreatment to post-treatment are in danger of relapse (Craig & Andrews, 1985; Craig et al., 1984).

Figure 8.12. Locus of Control of Behavior Scale.
Andrews and Craig (1988) reported that two measures of attitude, combined with a measure of stuttering behavior, are useful in predicting relapse after fluency-shaping treatment. They found little relapse among those stutterers who met the following three goals by the end of treatment: (1) no stuttering on telephone calls to strangers; (2) a score of 9 or below on the Modified Erickson Scale of Communication Attitudes; and (3) locus of control score reductions greater than 5 percent. Their assessment of relapse was based on a single telephone call with a stranger 10 to 18 months after treatment, and relapse was considered to be more than 2 percent of syllables stuttered during the call.
As mentioned in the preceding section, the OASES (Fig. 8.7) is an assessment tool based on the WHO’s 2001 International Classification of Functioning, Disability, and Health. It is a 20-minute paper-and-pencil questionnaire designed to obtain information about the impact of stuttering on a client’s life that is not usually gained by other measures. It is divided into four sections: I. General Information, II. Reactions to Stuttering, III. Communication in Daily Situations, and IV. Quality of Life. An “impact score” for each section as well as a “total impact score” can be calculated and then related to normative data so that a clinician can find out how severely stuttering is impacting a client’s life.
The developers of the OASES published data on validity and reliability (Yaruss & Quesal, 2006) on the adult version, indicating that internal reliability within each of the four sections was high (Cronbach’s alpha coefficient ranged from 0.92 to 0.97), suggesting that questions within a section were tapping into a homogenous area. In addition, Pearson product-moment correlations between total scores for these four sections were low enough (0.66 to 0.85) to indicate that the different sections were measuring different domains. Criterion-related validity was assessed by comparing an earlier version of the OASES to the Erickson S-24 Scale of Communication Attitudes (Andrews & Cutler, 1974; Erickson, 1969). Correlations suggested one section (Reactions to Stuttering) was positively and highly correlated with the S-24, whereas the other two sections (Communication in Daily Situations and Quality of Life) were moderately correlated. Test-retest reliability was assessed by giving the OASES to 14 individuals two different times, separated by 10 to 14 days without intervening treatment. Responses were identical for 77 percent of responses and within ± 1 for 98 percent of responses, suggesting that it is appropriate to administer the instrument repeatedly to an individual (e.g., before, during, and after treatment).
A summary of instruments to assess feelings and attitudes is given in Table 8.6.
Table 8.6 Methods of Assessing Attitudes and Feelings

Assessment is an ongoing process. As treatment progresses, the clinician should continue to ask herself, “Am I using the best approach with this person? Is there something else or something different I should be doing?” She should also decide what measures of progress are important for a client and apply these measures at regular intervals. My own approach is to assess stuttering behavior at the beginning and the end of each semester. In other settings, I often assess a client after every 10 hours of treatment. Although I don’t always succeed in these periodic assessments, I try to obtain samples of my clients’ speech in nonclinical situations, such as in the classroom or at work. I also do continuing assessment when I bring them in for maintenance checkups at increasingly longer intervals after formal treatment is over.
In addition, I assess clients’ feelings and attitudes at the beginning and end of treatment and may assess them at other times if I am concerned about progress. If I am working on changing attitudes and feelings, change should be reflected in my measures, or I should try a different approach. Decreases in negative attitudes and feelings should be accompanied by decreases in stuttering severity, and my measures should show this.

• In an assessment, the clinician has a variety of tasks. These include
1. Gathering data from the client
2. Getting to know the client as an individual
3. Showing an understanding of the client’s point of view
4. Demonstrating an understanding of stuttering
• Clinicians must develop skills and sensitivities in working with clients from culturesother than their own.
• Building a relationship with a client begins in the first meeting, which is usually theassessment. A clinician must take this opportunity to demonstrate that she knows about the disorder of stuttering, is unafraid of it, and is accepting of it. At the same time, she must show that she expects change.
• Behaviors counted as stutters include part-word repetitions, single-syllable whole-word repetitions, prolongations, blocks, and unequivocal avoidance behaviors.
• Samples for assessing stuttering should include a variety of situations. The initialsample and samples assessing outcome should be videotaped for more accurate scoring and measurement of reliability. These samples should include speaking and when appropriate, reading.
• Frequency of stuttering is commonly assessed as the percentage of syllables or words stuttered.
• Different types of disfluencies can be assessed to reveal the percentage of stutter-like disfluencies or number of these disfluencies per 100 words. This information may be particularly useful in helping to decide if a preschool child needs treatment.
• Durations of moments of stuttering are useful in quantifying an aspect of the abnormality of a client’s stuttering and the extent to which it may interfere with communication. Speaking and reading rates will also help to quantify this aspect of the impact of stuttering.
• Frequency and severity of secondary or concomitant behaviors associated with stuttering can be important measures of how much these behaviors call attention to themselves and distract listeners.
• Three severity scales are the Stuttering Severity Instrument (SSI-4), the Scale forRating Severity, and the Lidcombe Program’s Severity Rating Scale. The commonly used SSI-4 combines an assessment of frequency of stuttering, mean duration of the three longest stutters, and physical concomitants accompanying stuttering.
• Speech naturalness can be reliably and easily assessed. It is thought to be an especially important measure when evaluating treatment outcomes.
• Various instruments have been created to assess emotions and attitudes associatedwith stuttering. When combined with measures of stuttering behaviors, these measures provide a multidimensional view of the disorder. In some cases, they can aid the clinician in selecting appropriate treatment and in assessing whether a client is ready for dismissal.
• Assessment is an ongoing activity. Measures of progress are important indicators ofwhether ongoing treatment is effective and should be continued. Measures of outcome are critical for our knowledge of how effective a treatment has been in the long term with each client.

1. What does it mean to suggest that a clinician must play two different roles duringan evaluation?
2. Which types of disfluencies are counted as stutters, and which are considerednormal?
3. What factors can affect the process of measuring stuttering?
4. Describe the procedures for the two measures of reliability called “point-by-pointagreement” and “percent error.”
5. Why is it important to obtain several samples of speech from a client when assessing stuttering, whereas a single sample might be adequate for assessing a phonological disorder?
6. Give two reasons why assessment of reliability of measurements of stuttering isimportant.
7. Describe five dimensions or aspects of stuttering that may be assessed in an evaluation of a client who stutters.
8. To you, what is the most important aspect of stuttering? Defend your choice. Howdo you measure it?
9. Why is it relevant to assess speech rate for a person who stutters?
10. Discuss the pros and cons of protecting the privacy of conversations betweenyourself and a teenage client by not sharing their content with parents.
1. Research how different cultures react to stuttering and suggest how evaluationprocedures in this chapter need to be changed for individuals from a culture that has very different beliefs about stuttering than those suggested in this text.
2. Make or obtain a video of a conversational and a reading sample of a person
who stutters. Use two different methods of measuring the stuttering and obtain intraobserver and interobserver reliability assessments for each method. Discuss why one measurement procedure is more reliable than the other.
3. Obtain a videotaped conversational sample of one or more individuals whostutter and identify moments when you think the client has used an avoidance behavior to prevent stuttering. Discuss whether these avoidances should be counted as stutters.
4. Obtain samples of repetitive stutters from normally fluent children and fromchildren who stutter. Compare the repetitions from both groups in terms of length of silent periods between iterations, pitch, and other acoustic variables.
5. Search the literature on assessment of stuttering to determine whether reliablemethods have been developed for clients to assess their own stuttering during the progress of treatment.
Brutten, G., & Vanryckeghem, M. (2007). Behavior Assessment Battery for school-age children who stutter. San Diego, CA: Plural Publishing.
Conture, E. (2001). Assessment and evaluation. In Stuttering: Its Nature, Diagnosis, and Treatment. Boston, MA: Allyn & Bacon.
This chapter is rich with ideas for assessment of children, adolescents, and adults who stutter. It is particularly good in the breadth of coverage of evaluation of children who stutter, reflecting the years of experience the author has had in working with this age group.
Cordes, A. K. (1994). The reliability of observational data: I. Theories and methods for speech-language pathology. Journal of Speech and Hearing Research, 37, 264–278.
This article provides an excellent tutorial on the problems associated with establishing reliability of observational data in stuttering.
Howell, P., & Van Borsel, J. (2011). Multilingual aspects of fluency disorders. Tonawanda, NY: Multilingual Matters.
An edited book that discusses stuttering in many languages and cultures as well as stuttering and bilingualism.
Rosenberry-McKibbin, C. (2002). Multicultural students with special language needs: Practical strategies for assessment and intervention (2nd ed.).
Oceanside, CA: Academic Communication Associates, Inc.
This book provides many insights into evaluation of multicultural students.
Wright, L., & Ayre, A. (2000). The Wright Ayre stuttering self-rating profile. United Kingdom: Winslow Press Ltd. (
This assessment tool was designed to be used by clinicians working with adolescent and adult clients who can participate in assessing their own behaviors

and feelings. It is intended to go beyond traditional clinician-based assessments to obtain indications of how clients observe themselves.

Assessment and Diagnosis

Preschool Child
Clinical Questions
Initial Contact
Case History Form
Audio/Video Recording
Parent-Child Interaction
Parent Interview
Clinician-Child Interaction
Talking about Stuttering
A Child Who Won’t Talk
A Child Who Is Entirely Fluent
Speech Sample
Pattern of Disfluencies
Stuttering Severity Instrument (SSI-4)
Test of Childhood Stuttering
Speech Rate
Feelings and Attitudes
Other Speech and Language Behaviors
Other Factors
Physical Development
Cognitive Development
Social-Emotional Development
Speech and Language Environment
Determining Developmental and Treatment Level

Typical Disfluency
Borderline Stuttering
Beginning Stuttering
Risk Factors for Persistent Stuttering
Drawing the Information Together
Closing Interview: Recommendations and Follow-Up
Recommendations for Children with Typical Disfluency
Recommendations for Children with Borderline or Beginning Stuttering
School-Age Child
Clinical Questions
Public School Considerations
Initial Contact with Parents
Case History Form
Audio/Video Recording
Parent Interview
Teacher Interview
Classroom Observation
Child/Student Interview
Speech Samples
Pattern of Disfluencies
Stuttering Severity Instrument (SSI-4) Speech Rate
Trial Therapy
Feelings and Attitudes
Other Speech and Language Disorders
Other Factors
Physical Development
Cognitive Development
Social-Emotional Development
Academic Adjustment
Closing Interview
Public School Setting
Adolescents and Adults
Clinic versus School Assessment
Case History Form
Attitude Questionnaires
Audio/Video Recording
Speech Sample
Stuttering Severity
Other Measures of Stuttering
Speech Rate
Pattern of Disfluencies
Trial Therapy
Feelings and Attitudes
Other Speech and Language Behaviors
Other Factors
Interview with Parents of Adolescent
Determining Developmental and Treatment Level
Intermediate Stuttering
Advanced Stuttering
Closing Interview
After studying this chapter, readers should be able to:
• Plan and carry out an evaluation of a preschool child, school-age child, adolescent, and adult
• Understand how to evaluate the stuttering behaviors of a client
• Understand how to evaluate attitudes and feelings of a client (and when appropriate, the client’s family)
• Understand how to determine appropriate follow-up to evaluation of each agelevel
Preassessment: The time before the formal assessment, in which the clinician gathers key information needed for the assessment
Parent-child interaction: Play-based conversation between parent(s) and child that is often used to assess and treat environmental influences on a child’s stuttering
Parent interview: A conversation with parent(s) in which parents are encouraged to describe their child’s stuttering and their concerns about it. In this interaction, the clinician elicits and gives information relevant to the child’s stuttering
Clinician-child interaction: Play-based conversation between clinician and child in which clinician observes child’s speech and may seek information about child’s feelings about his stuttering, as well as information about child’s articulation and language skills
Speech sample: A segment of speech used to assess a client’s stuttering. A sample may be limited to 300 syllables. It may be spontaneous conversation or written work that is read aloud. It is meant to be representative of a client’s speech in general
Pattern of disfluencies: The types of stuttering and/or typical disfluencies shown by someone who stutters. Examples are whole-word repetitions, part-word repetitions, and prolongations
Risk factors for persistent stuttering: Characteristics within the child or within the environment that are hypothesized to increase probability that child will not overcome his stuttering naturally, that is, without intervention
Closing interview: A conversation at the end of an evaluation session in which the clinician informs the client or family of her findings, makes recommendations for the future, and elicits questions
Individuals with Disabilities Education Act (IDEA ’97): A federal law that mandates the procedures for gathering information and deciding on treatment of children in public schools
Teacher interview: A conversation with a child’s classroom teacher(s) for the purpose of getting information about the child’s speech and performance in the classroom. Such a conversation may also be used to enlist the teacher’s help with the child’s treatment
Classroom observation: Time spent in the classroom by the clinician to assess how a child’s stuttering may be affecting him in his classes
Accepting environment: Behavior by parents and other family members (and teachers and classmates where appropriate) that convey to the child that they accept his speech
Trial therapy: A brief administration of one or more therapy strategies, used for the purpose of determining the effect on the client’s speech in an evaluation
Fluency skills: Ways of speaking designed to induce fluency. Examples are slow rate, easy onset of voicing, and light contact of articulators
Individualized Education Program (IEP): A plan, mandated by the Individuals with Disabilities Education Act, that describes how a person with a disability will be educated to meet his individual needs
Stuttering modification: Ways of managing stuttering that are designed to reduce struggle and tension. Examples are cancellations, pull-outs, and preparatory sets
This chapter and the preceding one are bridges between chapters on the nature of stuttering and chapters on treatment. My aim is to show you how to understand a client and his stuttering problem and then use the information you have gathered to determine a treatment approach. Figure 9.1 has a chapter overview, and Figure 9.2 illustrates the components of assessment and diagnosis.

Figure 9.1 Chapter overview.

Figure 9.2 Sequence of assessment and diagnosis.
I’ve organized the assessment procedures in this chapter by age levels: preschool, school-age, and adolescent/adult. I’ve done this in part because in preparing to evaluate a client, you often know little about him except his age, so you will at least know what section of this chapter to turn to as you plan an evaluation. Another reason for this organization is because each age level requires a different approach and somewhat different procedures. When evaluating preschool children, for example, the clinician must determine whether they are stuttering or normally disfluent and then whether treatment, if warranted, should focus on the family or on the child’s speech. With school-age children, a clinician often follows procedures that will allow her to develop an Individualized Education Program (IEP) for children who are eligible for services. With adolescents and adults, a clinician may more extensively assess emotions and attitudes as well as core and secondary behaviors to determine appropriate treatment.
Within each age level, I have subdivided the sequence of activities you will typically engage in when you evaluate a client. First, under the preassessment sections, I describe the clinical questions guiding the evaluation, as well as preliminary information-gathering activities. Under the assessment sections, I have described the observations, interviews, and measurements at the heart of the evaluation. Under the diagnosis sections, I discuss how to integrate the information you have gathered to decide on a course of action (whether to recommend treatment and if so, the type of treatment). Last, under the Summary and Recommendations sections, I describe the closing interviews, assignments, and the means to assess progress and outcome.
Clinical Questions
When you assess a preschool child, you first want to answer the question, “Is this child stuttering, or is he normally disfluent?” In the process of trying to answer this, I have found that certain other information is critical: What are the amounts and types of disfluency this child shows in various situations? What kinds of risk factors for stuttering does the child have? Is the child reacting to his speech with frustration, fear, or other emotions? This information will also help you determine the developmental/treatment level of the child’s disfluency. To respond to the family’s needs and to involve them fully in planning treatment, you need to know the answers to questions like: What are the family’s concerns about the child and his stuttering, as well as their preferences, expectations, and availability for treatment? Once you have gathered this information, you will need to decide among options for the child and family–no treatment, watchful waiting, parent-delivered treatment, or clinic-based treatment. You will also want to decide if the child is within the normal range for language, articulation, and voice. Finally, you should determine if a referral to another professional (e.g., a family counselor or a learning specialist) is warranted.
Initial Contact
In your initial contact, the family member who has approached you is forming his or her first impression of you. In many cases, this is the beginning of a helping relationship. You should focus on understanding the other person’s point of view, concerns, and hopes. When you can, give realistic reassurance about the prospect of recovery, but in all cases, convey that you are ready to work with the family member as part of a team who will, together, help the child.
Your initial contact is likely to be on the telephone. If it is, listen to the parent’s voice and pay attention to the parent’s level of concern. Often, the most helpful thing you can do in this conversation is to listen. As you listen carefully, you may need to ask an occasional question to get clarification. When you think you have an initial understanding of the problem, set up an appointment, if appropriate. If you cannot meet with the family for several days or longer, it is important to give them some suggestions to get started. For example, you may want a parent to rate the child’s stuttering using the Severity Rating Scale described in Chapter 8. It also may be helpful to have the family set aside a few minutes every day, in which one parent can play alone with the child and give him special attention. Also, consider letting the family know that they can contact you before the first formal meeting if they have new concerns. Prior to an evaluation, I let the family know what will take place in the evaluation and approximately how long it will take. A discussion of fees and payment may also be appropriate.
Case History Form
The case history form, shown in Figure 9.3, is sent to parents several weeks before their child’s assessment. It informs the clinician about the parents’ current perception of the problem, as well as its onset and development, and the child’s medical, family, and school history. This information is used as a starting place for further questions during the parent interview. More about this will be discussed when the actual evaluation is described.

Figure 9.3 Case history from preschool and school-age child
Audio/Video Recording
Along with the case history form, I ask the parents of a preschool child to send me an audio recording or preferably a video recording of their child speaking in a typical home situation. I encourage parents to video five or 10 minutes of themselves playing with their child so I can preview the child’s speech soon after the parents have contacted me. In cases in which several weeks go by between the parents’ contact and the evaluation, the child’s stuttering may have diminished substantially, and I may observe only a fluent cycle of the child’s speech. In addition, I can learn from the recording a little about this family’s parent-child interactions.
Parent-Child Interaction
When possible, I observe one or both parents interacting with the child, preferably at the beginning of the evaluation for several reasons. First, parents may be less affected by my orientation toward stuttering, which they would learn during the parent interview. So observing the parent-child interaction first gives me a more natural sample. Second, this interaction gives me an opportunity to see the child’s stuttering firsthand. I can note, for example, how much the child seems aware of his stuttering, whether or not there are escape and avoidance reactions, and to what extent the child is reacting emotionally to his stuttering. Third, I can observe how the parents interact with their child. Do they interrupt? Do they correct? Do they talk at a fast rate or use complex vocabulary or advanced syntax? Fourth, if I began the evaluation by talking with the child, he might not talk much, since he doesn’t know me. Thus the parentchild interaction will give me at least a larger sample of his speech. Then when I interact with him, I can get a sample of how his speech is with new people. Such observations of the parent-child interaction add to what I have learned from the audiovideo sample that the parents have sent before the assessment. Together, the observation of home and live samples provide a basis for planning the parent interview and developing recommendations for treatment.
The parent-child interaction can be done formally or informally. Some clinicians observe these interactions in the waiting room. Others who work in preschool or early intervention programs may visit a child’s home and arrange to observe parent-child interactions while they sit quietly in the same room. Still others, myself included, videotape the parents and child in a play-style interaction in a treatment room supplied with toys and games. When recording these interactions is possible, this sample of a child’s speech can be assessed for severity and types of stuttering behavior, as described in Chapter 8.
Parent Interview
In the last few years, I have talked more freely about stuttering directly with preschoolers, as well as with their parents when their child is within earshot. I think this directness may reduce parents’ distress about their child’s stuttering, perhaps by reducing everyone’s fear of it. When I first meet a child and his parents, however, I hold back a little until I have had a chance to discuss this openness with the parents. The child may be only normally disfluent, or the parents may be reluctant to talk about stuttering in front of the child. The parent interview gives me a chance to suggest how, in my experience, openness about stuttering decreases some of the shame a child (and parents) may feel about stuttering. First, I talk to parents without the child present, giving them an opportunity to talk about matters that they feel they would like to share in confidence. If I’m working with another clinician or a student, she plays with and observes the child while I talk to the parents. If I’m working alone and both parents have come to the evaluation, I talk to each parent separately, while the other is with the child. If only one parent is present, I often arrange to have the child playing by himself in a nearby room with the door open.
I begin the interview by letting parents know what I will be doing with them and their child during the remainder of the evaluation, even though I may have touched on this in the first contact. I assure them there will be a time for me to share my assessment and recommendations with them at the end. Sometimes during an initial interview, parents ask direct questions about things they think they may be doing wrong. I let them know that, in my view, stuttering is often the result of many factors acting together and that parents do not cause it (see Chapters 2 to 5 for evidence). I rarely give advice about what they should change or what they should do until after I have finished interviewing the parents and assessed the child directly. I believe that I am more accurate, and parents are more receptive to my recommendations if I delay a discussion of what to do until the closing interview when I have the most information possible. On the other hand, there may be “clinical moments” during an initial interview when parents might be most receptive to suggestions. Many times, for example, parents have asked me whether it’s a good idea for them to tell their child to slow down whenever the child stutters when excited. My response is usually to ask them how the child responds to this and to build upon their answer so that we can brainstorm the best way to help the child together.
I begin initial interviews by asking parents to describe the problem their child is having. I ask open-ended questions, such as “Tell me about Justin’s speech,” or “Please describe Justin’s speech and tell me what concerns you.” Open-ended questions allow parents to describe their concerns in their own words. This is then an opportunity for me to listen carefully, to be nonjudgmental, and to be comfortable with silence so that parents can express themselves fully. Listening attentively and being comfortable with silence require concentration on the clinician’s part, so it is worthwhile to remind yourself of this when you are preparing for an evaluation. When parents have had a chance to describe the problem and appear to have no more to say at that moment, I ask about the first stages of the child’s life (the child’s birth and development) and then work up toward the present time. In the ensuing conversation, I try to be sure I get information indicated by the questions in the following paragraphs. This interview is not a strict question-and-answer format, but rather a discussion punctuated by both their questions and mine.
1. Were there any problems with your pregnancy or the birth of this child? Although there is little evidence that stutterers as a group have difficult birth histories, there is an increased incidence of stuttering among individuals who have a known history of brain injury (Boehme, 1968; Poulos & Webster, 1991). Thus, I am seeking to determine whether there is the possibility of congenital brain injury. If a difficult pregnancy or birth is reported, I might examine the child’s motor and cognitive development more closely. Because it is completed before the evaluation, the case history form provides preliminary information for potential follow-up.
2. What was the child’s speech and language development like? How did it compare with siblings’ development and with your expectations?
The first appearance of stuttering may be influenced by the “processing load” that language acquisition has on a child’s speech production, as described in the sections on speech and language development in Chapters 4 and 5. Thus, it is important to understand the course of a child’s overall speech and language development. I explore the possibility that a child’s language acquisition is proceeding so rapidly that his developing motor system cannot keep up. I also examine the possibility that a child’s speech and language development are delayed, and he is frustrated and finding it hard to talk. As mentioned in Chapters 2 and 3, there is some evidence that poorer speech and language skills may predict persistent stuttering (Yairi et al., 1996).
3. Describe the child’s motor development compared with that of his brothers, sisters, or with other children.
I am interested in parents’ general impressions. Does their child seem to be developing motor skills like other children his age, or do they think he may be delayed? Some indicators of the normal range of children’s gross and fine motor development, as well as their personal-social and speech-language development, can be found in the Denver Developmental Screening Test (Frankenburg & Dodds, 1967) and is shown in Figure 3.2 (child development in the first five years).
In my experience, many children who stutter appear to be slightly advanced in their language and to a lesser extent, slightly delayed in their motor skills. Or, they may be well advanced in language but with completely normal motor skills. In either case, these children seem to benefit from models of speech produced at a slow rate (Guitar, Kopff-Schaefer, Donahue-Kilburg, & Bond, 1992; also see the section in Chapter 11 on indirect treatment). Other children who stutter may be delayed in several areas and may need treatment for language and articulation that is integrated with therapy for stuttering (see the section of Chapter 12 on Treatment of Concomitant Speech and Language Problems).
4. Have any members of your family had speech or language disorders? I ask this general question and then ask more specifically whether family members or other relatives have ever had problems related to stuttering, cluttering, articulation, or language disorders (see Chapter 15 for a description of cluttering). To confirm that a problem was considered significant (and was perhaps diagnosed), I ask if the person ever received treatment. I use this information when we discuss stuttering as a disorder that may have predisposing factors. Handled tactfully, a discussion of predisposing factors may help parents realize that their child’s stuttering was not something they caused, which in turn may reduce their anxiety or guilt, making them more effective in facilitating the child’s fluency.
If a parent stutters or formerly stuttered the parent may have strong negative feelings about the disorder, including guilt that the parent has passed it on to the child. Such feelings should be discussed in the initial interview and throughout any treatment the child receives. The way in which a parent who stutters handles his or her stuttering is also important, because these behaviors serve as a model for the child. It is my observation that a parent who avoids words or otherwise tries to hide stuttering is communicating an attitude that may move the child to the intermediate level faster than if the parent accepts the stuttering, comments neutrally about it in front of the child, and uses facilitating techniques to handle it.
If any of the child’s relatives stutter, it is important to find out whether they recovered. Research cited earlier found that among children who were identified within six months of the onset of stuttering, those with relatives who did not recover from stuttering were more likely to have persistent stuttering than those with relatives who did recover (Yairi et al., 1996). After obtaining this background information, I turn to the onset and development of the child’s stuttering.
5. When did you first notice the child’s disfluency?
I have found that if treatment is begun relatively soon after a child starts to stutter(within 18 months, rather than after several years) we have a better chance of preventing negative feelings from building up for both the parents and the child. Therefore, I praise parents for bringing a child in promptly for an evaluation, if they did so relatively soon after they first realized there may be a problem. Another reason I want to know how much time has passed since onset is that most of the predictive information on chronicity of stuttering is based on children identified within six months of onset. For example, Yairi and colleagues (1996) found that children who naturally recovered began to show a steady decline in their stuttering during the first 12 months after stuttering onset, whereas children whose stuttering persisted for at least three years did not show such a decline. Therefore, knowing how long a child has been stuttering helps me make treatment decisions based on findings that some children are likely to recover without therapy.
6. Was anything special going on in the child’s life when the stuttering started? This may provide some leads about the kinds of pressures to which a child may be vulnerable, which can help clinicians determine what changes parents can make to reduce stuttering. Events that may precipitate the onset of stuttering include the birth of a sibling, moving to a new home, family travel, prolonged periods of anxiety or excitement, and growth spurts in a child’s language or cognition (see Chapters 4 and 5). Many times, no special circumstances have occurred at the onset of stuttering. Events surrounding the onset of stuttering should be discussed in a way that helps parents feel they are not to blame for the stuttering.
7. What was the disfluency like when it was first noticed?
Most stuttering begins with easy repetitions, although some children exhibit prolongations and blocks, as well. Some preliminary information suggests that when repetitions sound quite rapid (i.e., when the pause between repetition units is brief), a child is more likely to be stuttering rather than normally disfluent (Allen, 1988; Throneburg & Yairi, 1994). Rapid-sounding repetitions may be predictive of persistent stuttering (Yairi et al., 1996). However, the length of pauses between repetition units cannot be determined accurately without instrumentation, even though a practiced ear can help clinicians perceive the brevity of pauses between repetition units. This information should be used only to support an overall pattern of findings that will help the clinician decide whether or not to recommend treatment.
8. What changes, if any, have been observed in the child’s speech since stuttering was first noticed?
The most interesting changes include the frequency and types of disfluencies and whether and for how long the stuttering diminished greatly or disappeared altogether. As indicated in the discussion of Question 5, children whose frequency of core stuttering behaviors (i.e., part-word and single-syllable wholeword repetitions, prolongations, and blocks) does not decrease during the 12 months after onset are at risk for becoming persistent stutterers. In my clinical experience, if a child’s physical tension and struggle during stuttering are increasing, or if stuttering is becoming more consistent and less intermittent, the child is not exhibiting a borderline level of stuttering, and direct treatment should be considered.
9. Does the child appear to be aware of his disfluency?
If a child appears to have no awareness of his disfluencies, I am more likely to categorize him as normally disfluent or as a borderline stutterer than if he notices or seems concerned about his disfluencies. If he shows negative awareness, such as expressing frustration, he may be a beginning stutterer. Note that a child may be aware of his stuttering but not particularly bothered by it; some children are even amused by it when it first occurs, though this quickly changes to frustration. Indicators of a child’s awareness include such things as his commenting about his stuttering, either when it occurs or at some other time, and responses to the fact that people have brought it to his attention. Awareness is also indicated if a child stops when he is disfluent and starts again or laughs, cries, or hits himself when he stutters. Even without any of these signs, a child may still be quite aware of his stuttering.
In some cases, preschool children may show more than just signs of frustration. They may show negative feelings about talking and may fear using certain words. They may even comment that they wish they could speak like someone else. These signs of awareness are indications that treatment is warranted.
10. Does the child sometimes appear to change a word because he expects to bedisfluent on it?
Parents are usually able to guess this is happening because they can sense the child’s apprehension about saying a word. I also may ask them if the child changes words in midstream; that is, does he start a word, get stuck on it, and then change it or stop talking? Such behaviors are warning signs; they may indicate that the child is moving toward a more serious problem.
11. Does the child seem to avoid talking in some situations when he expects to be disfluent?
Again, this is something that most parents know because they sense the child’s fear of talking, and like the word avoidances discussed in Question 10, this behavior may indicate a need for direct treatment.
12. What do the parents believe caused the problem?
In some cases, parents may express ideas about the possible causes of their child’s stuttering that I believe are appropriate and accurate. In other cases, parents’ beliefs about causal factors appear to be incorrect, and I respond by providing more accurate information. I am particularly sensitive to whether or not parents blame themselves or each other for their child’s stuttering. This is usually a good time to let parents know that they are not to blame. I tell them that some children may have slight differences in their neurological organization for speech, which may emerge as stuttering during the normal stresses and strains of growing up and learning to talk (see Chapters 2 to 5). Parents should know that they didn’t cause their child’s stuttering, but they should also know that they can play a key role in their child’s learning to deal with it appropriately.
13. How do the parents feel about the child’s disfluency problem?
The kinds of feelings and attitudes we are looking for are: Do they feel concern? Guilt? Do they assume the child will outgrow it? Parental emotions and attitudes are contagious and may influence the child, particularly a sensitive child. If parents feel guilty or highly anxious, it is important to engage them in positive treatment activities as soon as possible. I suggest some beginning activities they can use in the Summary and Recommendations sections.
14. What, if anything, have the parents done about the child’s disfluency problem?
This question is aimed at finding out how the parents have responded to the child’s disfluencies. For example, have they asked the child to slow down or stop and say the word again? Knowing this will help me to decide what to do in counseling them. If parents are correcting the child, I may get them involved in therapeutic activities immediately so that they can develop appropriate ways of responding.
15. Has the child been seen elsewhere for the problem? If so, what were theoutcomes?
This information can be important in planning therapy and counseling parents. For example, if their family doctor told them several years ago that the child will outgrow stuttering, this needs to be addressed because they may now be convinced he will not outgrow it. It is wise to comment positively or neutrally on what other professionals may have said or done. So many children appear to overcome stuttering without treatment that most doctors and nurses believe that their advice to parents to ignore the stuttering will have a good outcome. Some doctors and nurses, however, are learning to distinguish between children who are likely to recover without treatment and those who are not.
If the child has been in other treatment previously, knowing what advice the parents were given can be important. Sometimes, parents have been given excellent advice but were not able to follow it. If so, we need to find out why and help them overcome obstacles to helping their child. Sometimes parents have had their child in successful therapy but have moved away and sought me out to continue the same kind of treatment. In these cases, I try to contact the previous therapist as well as explore with the parents what was done so that we can continue to work in the same direction as before. In some cases, parents come to me seeking a second opinion, and I am able to reinforce what others have said if I agree. In other cases, they may have been advised to ignore the child’s stuttering, which may lead me to tactfully discuss the possibility of taking an entirely different direction.
16. When and in which situations does the child exhibit the most disfluency? The least disfluency? This information helps to identify fluency disrupters and fluency facilitators that I will use to help parents facilitate their child’s fluency. I have also found it effective to point out whenever possible all of the helpful things the parents are already doing. Just the awareness that their child’s stuttering responds to environmental cues and thereby has some logic to it helps most parents feel more competent to manage it.
17. How does the child get along with his brothers and sisters and other children?
I usually find that children who stutter relate fairly well to others, but I want tofind out if a child’s stuttering may be interfering with his relationships. Sometimes, when asking this question, I learn about pressure and competition from siblings or teasing by a neighborhood bully.
18. What is the child’s temperament like?
Some children who stutter may be more emotionally reactive than most other children, and they may have less capacity for self-regulation (i.e., dealing with that reactivity). More emotionally reactive children with less ability to selfregulate would be more likely to respond negatively to parents’ anxieties about their speech. A child with this temperament may benefit from extra help in learning to cope with arousing stimuli. There is good evidence that families can help a child develop a more resilient temperament (Calkins & Fox, 1994).
19. What is a typical day like for your child?
It can be helpful to get an idea of how busy and rushed a family is. For one thing, it has been my experience that many children who stutter and their families benefit from having less hectic schedules, particularly if the child thrives in more regular, less intense situations. You can add this information to what the parents told you about when their child stutters most to develop a hypothesis about how much the family’s schedule may be affecting the child’s fluency. If the child stutters more when things are busy, hectic, and stressed, it may be appropriate to brainstorm with the parents about how everyone can have a little more “down time.” Knowledge of the family’s schedule will also help you begin to consider treatment recommendations. Some treatments are demanding of parents’ time and attention, and their schedule must be considered in working with them to determine the most appropriate treatment approach for their child.
20. Is there anything else you can think of to tell me that will help me betterunderstand your child’s stuttering?
Sometimes, it is not possible to direct questions to all areas of concern, and this question provides parents an opportunity to provide information that I have not thought of asking about.
Clinician-Child Interaction
One of the most important parts of a preschool child’s evaluation is the clinician-child interaction. Here, the clinician can see firsthand what the child’s speech is like, how he responds to various cues, and how well he can modify his disfluency. I always record this interaction for later analysis because it is difficult to make notes as we interact. Video recording is preferable because visual cues are often critical in determining a child’s developmental and treatment level. If audio recording must be used, the clinician should make notes on visual aspects of the child’s disfluencies.
I focus my interactions on toys or games that are suitable to the child’s age. The
Playskool® farm or airport is a good example. I play alongside the child, letting him direct the action, commenting on what he’s doing or playing with. I refrain from questions as we begin and talk in an easy, relaxed manner, much as I advise parents to do.
If a child’s stuttering is like that described by the parent, I maintain the same speech style throughout the interaction. However, if a child is entirely fluent or normally disfluent and the parents have described behaviors typical of stuttering, I speak more rapidly and ask many questions. Occasionally, I interrupt the child to elicit disfluent speech, which may be more characteristic. I do this to avoid misdiagnosing a child who is stuttering as a normally fluent speaker.
An adult client of mine described an experience that illustrates my concern. When she was 5 years old, she stuttered quite severely, and her parents were understandably concerned. Seeking the best help, her mother took her to a famous Midwestern university speech clinic for an evaluation. For reasons she never understood, she was relatively fluent throughout the entire evaluation. The clinicians observed her temporary fluency and despite her mother’s protestations that her daughter stuttered at home, did not diagnose her stuttering and advised her mother to ignore any disfluency. Her disfluency gradually worsened, and she became a severe, chronic stutterer.
I realize that, even by putting pressure on the child, I may not elicit stuttering that the child displays in other settings. Thus, the parents’ report and the recording they made before the evaluation are of vital importance for a full understanding of a child’s speech.
Talking about Stuttering
Before interacting with a child, I try to determine from talking with a child’s parents whether the child is aware of his stuttering. If I think he isn’t, I simply observe the child’s speech while he and I play. If it seems clear that he is aware of his stuttering, I then try to determine how comfortable the child is in talking about his stuttering. Sometimes, I ask if he knows why he has come to see me. Most children answer noncommittally, but some say something like, “Because I don’t talk right.” This gives me an opening to discuss the stuttering. It is also an important opportunity to let the child know that he isn’t alone and that I know other children who get stuck on words and am usually able to help them.
Some clinicians help a child talk about his stuttering by first talking about another child who stutters (Bloodstein, personal communication, 1990). In discussing stuttering with a child, I usually try to use their vocabulary, such as “getting stuck” or “having trouble on words.” If a child seems reluctant to talk about stuttering, I drop the issue for the moment and return to playing. Then, later, I will insert a few naturalsounding disfluencies in my speech and comment that I sometimes have trouble getting words out. I might play some more and then insert a few more disfluencies and ask the child if he ever has trouble like this. As before, the child’s response will indicate that either he remains unwilling to discuss stuttering or he will give the clinician an opening to discuss, little by little, his disfluency problem. In summary, the goals of these attempts to discuss a child’s disfluency are (1) to see if he accepts himself and his disfluencies enough to discuss them, and (2) to assure him that he is not alone with the problem and that his parents and I may be able to help him.
A Child Who Won’t Talk
At times, I encounter a preschooler who is reluctant to separate from his parents. A shy child may start to cry and cling to his parents. I don’t force the child to separate, of course. It is more important to have him positively inclined toward therapy than to try to elicit a few stutters. In this situation, I sit quietly while the parent and child play together. After a few minutes, I join in the play, without focusing on the child, and after a few more minutes, I’ll comment on what the parent and child are doing or what I’m doing with a tractor or a farm animal or whatever I’m playing with. In most cases, the child will soon say something to me or include me in the play. This interaction, leading to at least a little speech from the child, gives me an opportunity to observe, at close range, any stuttering the child may have. Only after a child gets comfortable with me do I attempt to discuss his trouble talking, and only if I’m sure he is aware of his stuttering.
With some children, I do not attempt to discuss stuttering, and I always take my cue from the child and go slowly in this area. A very shy child, who becomes even shyer if I produce a few easy disfluencies, may be quite turned off to therapy if I invade his space by asking about his stuttering at this point. You can infer many things about a child’s feelings from observations rather than direct questions.
A Child Who Is Entirely Fluent
Some preschool children who stutter may be entirely fluent during an evaluation. In such cases, there are several options. First, the recording I asked the parent to send me may include enough stuttering to provide a good sample for analysis. Second, if a child is in a particularly fluent period, I may reschedule his evaluation for a later time. If my recommendations to the parents enable them to change the home environment enough in the meantime so that the child remains fluent, the parents may wish to postpone the evaluation until and if the child’s stuttering returns.
Speech Sample
The following sections describe how to analyze samples of a preschool child’s speech. You should have more than one speech sample to analyze from the recordings that the parents sent in, the parent-child interaction, and the clinician-child interaction. Because you may want to use the SSI-4 (Riley, 1994) as part of your assessment, you need to follow the procedures it recommends for this analysis. Thus, the sample obtained from the clinician-child interaction should include conversation using the pictures in the SSI-4. Riley recommends that as the child talks, the clinician should “interject questions, interruptions, and mild disagreements to simulate the pressures of normal conversation at home and elsewhere” (Riley, 1994, p. 7). The samples should include at least 200 syllables; samples this long or longer make it more likely that you will have an accurate picture of the child’s speech. By making transcripts of the samples, you can more easily quantify the variables described in the following section on pattern of disfluencies.
Pattern of Disfluencies
By analyzing the child’s speech sample, I can try to determine whether or not the child truly stutters and if so, his developmental/treatment level. I analyze the following six variables to begin this determination. The choice of variables owes much to four individuals who have written about the differential diagnosis of preschool stuttering (Adams, 1977; Curlee, 1984; Riley & Riley, 1979).
1. Frequency of disfluencies. This is calculated from the entire sample and is expressed as the number of disfluencies per 100 words (see Chapters 7 and 8 for details). Both normal disfluencies and those associated with stuttering are included in this count. Normally disfluent children usually have fewer than 10 disfluencies per 100 words.
2. Types of disfluencies. I described the following eight types of disfluencies in Chapter 7: part-word repetitions, single-syllable word repetitions, multisyllable word repetitions, phrase repetitions, interjections, revisions-incomplete phrases, prolongations, and tense pauses. Children who are normally disfluent are likely to have more revisions and multisyllable whole-word repetitions, as well as many interjections when they are younger than 3.5 years old. Part-word repetitions, single-syllable word repetitions, prolongations, and tense pauses occur more frequently in stuttering children. Another distinguishing measure is the proportion of total disfluencies that are stutter-like disfluencies (SLDs) (i.e., part-word repetitions and single-syllable repetitions, prolongations, and blocks). Less than half of the disfluencies of normally disfluent children are SLDs, but about twothirds of the disfluencies of children who stutter will be SLDs (Yairi, 1997a).
3. Nature of repetitions and prolongations. This variable has several dimensions. First,normally disfluent children usually have only one extra unit in their repetitions, lilike this, but sometimes they may have two. As the number of repetition units increase, however, so does the likelihood that the child is stuttering. Second, I listen to the tempo of repetitions. If they are slow and regular, a child is more likely to be categorized appropriately as a normally disfluent speaker. If they are rapid or irregular, it is more likely that the child is stuttering. Third, I look for signs of tension in both repetitions and prolongations. Both visual and auditory cues can help here; tension can be seen in the child’s facial expression and heard in his increased pitch or loudness and more staccato voice quality. Children who I would label as normally disfluent seldom exhibit tension in their disfluencies.
4. Starting and sustaining airflow and phonation. The child who we usually categorizeas stuttering often has difficulty here. You may observe abrupt onsets and offsets of words, especially repeated words, or momentary pauses with fixed articulator positions at the onset of words. Moreover, transitions between words may seem abrupt, jerky, or broken much of the time.
5. Physical concomitants. I look for physical gestures that accompany a child’s disfluencies, such as head nods, eye blinks, and hand or finger movements, especially gestures that coincide with the release of a disfluent sound. I also include such extra noises as a child gritting the teeth or clicking the tongue during disfluencies.
6. Word avoidances. Another sign I sometimes see in a disfluent preschool child,which suggests that he stutters, is word avoidance. This can be blatant, as when a child starts a word and then changes it, as in “pu-pu-pu … dog,” or it may be more subtle, as when he says, “I don’t know,” when it’s clear that he does know. I also ask about word avoidances when I interview a child’s parents. When a clinician interacts with a child, she may sometimes miss avoidances in a live interaction, and it may take a viewing of the videotape to pick them up. For example, a few years ago I noted on the videotape I watched after an evaluation a very subtle avoidance that I had completely missed during the face-to-face interaction. I had asked the child what he was going to dress as for Halloween. He pursed his lips for a “B,” but when he couldn’t say the word, he used an avoidance by singing the Batman theme, “Na-na-na-na-na-na-na-nah! Batman!”
In my experience, if a child shows any of the characteristics of stuttering just described, he should be considered at least a borderline stutterer. The presence of tension, stoppage of airflow or phonation, physical concomitants, or word avoidances would place him on a level above borderline. Further details on this placement are given in the sections on diagnosis that follow.
Stuttering Severity Instrument (SSI-4)
Using the 200-syllable or longer samples gathered earlier, you should carefully follow the guidelines in the examiner’s manual of the SSI-4 to determine a child’s stuttering frequency, duration, and physical concomitants scores. Use of the SSI-4 is described more fully in Chapter 8. Using the definitions of stuttering given by Riley (1994) in the manual, which essentially are what I referred to earlier as “stuttering-like disfluencies,” you can use the total overall score to derive a percentile ranking for the child, which compares him to the norms for children who stutter. In addition, rankings that range from very mild to very severe can also be derived from the total overall score. Sometimes, normally disfluent children may be rated as stuttering at the very mild level on the SSI-4. Thus, clinical judgment, informed by your analyses of the types and frequencies of disfluencies, must be used to sort out which children are actually stuttering and which are not. The SSI-4 is not a tool for differentiating stuttering from normal disfluency but for assessing a child’s severity.
Test of Childhood Stuttering
As described in Chapter 8, the Test of Childhood Stuttering (TOCS) can be used for children ages 4 to 12 and is thus appropriate for older preschool children. This instrument evaluates the child’s stuttering in a variety of speaking situations and provides a more in-depth analysis of the child’s stuttering, but takes more time to administer than the SSI-4. I have not used it enough to give an informed opinion about the comparative usefulness of the TOCS and the SSI-4.
Speech Rate
I assess preschool children’s speech rate using the speech sample obtained for the SSI4. Counting and timing procedures were described in the section on assessment of speech rate in Chapter 8. One sample of speech rates for preschool children is given in Table 8.5. If a child is stuttering and his speech rate is substantially below the range for his age, the extent to which stuttering slows his rate of speech may be a problem for both listeners and the child. Children whose rates are substantially above the norms –or who sound like they are talking too fast–may have the disorder of cluttering, which is described in more detail in Chapter 15.
Feelings and Attitudes
If I am evaluating a preschooler, I begin by asking the parents if they think the child is aware of his stuttering, and I explore observations they have made of his reactions to his disfluencies. If they are convinced that the child is oblivious to his disfluencies and I observe those disfluencies to be without much tension and struggle, I don’t talk directly to the child about his stuttering. Instead, I rely on what the parents can tell me during the parent interview about their child’s feelings. I am also beginning to use the Impact of Stuttering on Preschoolers and Parents questionnaire (see Chapter 8) to gather initial information about the child’s (and parents’) feelings and attitudes about stuttering. However, if the child shows signs of struggle and tension when he stutters, or if the parents indicate that the child is aware through various examples of his frustration with his stuttering, I explore with the child his feelings about getting stuck on words.
Before bringing up the topic of stuttering, I get to know the child by talking with him during various play activities. Then, as we play, I insert a question about his speech, such as “Do you sometimes get stuck on words?” Both the child’s verbal and nonverbal responses to a gently asked question about stuttering tell me a lot. Even beyond what I notice when we are talking, I am often able to learn a great deal by watching the video of my interaction with a child. I find that by playing a video recording of my interaction, I am able to devote my undivided attention to observing key segments of the interaction. The recording often provides a rich payload of information about a child’s feelings that may not have been apparent to me in the faceto-face meeting. Some children may be quite comfortable answering my question about getting stuck on words, while others are embarrassed, look away, and don’t make clear responses. Still others emphatically deny they have any problem talking. If the stuttering is very mild and the child matter-of-factly says he doesn’t get stuck, I may tentatively conclude that he really isn’t aware.
Assessment of the feelings and attitudes of a preschooler leads me to conclude tentatively whether a child (1) is unaware of his disfluencies, (2) is occasionally aware of them and even then, is seldom and only transiently bothered by them, (3) is aware and frustrated by them, or (4) is highly aware, frustrated, and afraid of them. The levels of awareness and emotion that a child has about his stuttering are an important consideration in planning treatment, as we shall see.
Other Speech and Language Behaviors
When I evaluate a preschool child’s speech for stuttering, I also screen for possible articulation, language, and voice problems. In addition, I make sure that his hearing has been checked recently and if not, arrange to have a hearing screening.
A child’s language and articulation problems can usually be detected in the recorded parent-child or clinician-child interactions. When I suspect problems in these areas, I administer formal tests. You may wish to consult Bernthal, Bankson, and Flipsen (2009) for testing articulatory and phonological disorders, and Paul and Norbury (2011) or McCauley and Fey (2006) for assessing language problems. I will discuss the management of concomitant articulation and language disorders in Chapter 12, which deals with treatment of beginning stuttering.
My view of the relationship between language and stuttering, which I described in Chapters 4 and 5, is that one of the pressures on a child who stutters may result from language development that is much more advanced than motor development. Thus, in evaluating a child’s language and articulation, I explore the possibility that his language exceeds age expectations. In addition, I observe his language usage and motor abilities and question parents about his general motor development and the intelligibility of his speech.
When language development outstrips motor development, there may be a risk that a child will try to produce long sentences at an adult pace with a speech system that at this age, is better suited to a slower rate. A child’s motivation to speak quickly may come from his own eagerness to express complex thoughts, from his parents’ pleasure at his adult-like speech, or just from the fact that adult speech rate models affect the child. We have demonstrated this with typically developing children–typical in both language and fluency (Guitar & Marchinkowski, 2001). For a child who stutters and who also has advanced expressive language abilities for his age, rate of speech production may be an important factor to target in treatment. How rate is targeted in intervention depends on the child’s level of stuttering. If the child is relatively unaware of his stuttering and does not seem to be reacting to it with escape or avoidance behaviors, and his frequency of stuttering is relatively low, I am likely to use an indirect treatment approach. I would train parents to use a slower speech rate when speaking to the child as part of the treatment, with the expectation that their model of a slower speaking rate will influence the child to speak more slowly, thereby putting fluency within his reach. In such cases, I also explore ways in which the family may be putting pressure inadvertently on the child’s language skills by expecting a higher level of language development than the child can achieve.
Verbal activities that some parents enjoy most with their children, such as puns, word play, and teaching the child multisyllable words, may convey to a child that the parents place high value on verbal ability. For most children, this would be an incentive to develop their verbal skills. But for children vulnerable to fluency breakdowns, their parents’ pride in their verbal proficiency may stress their ability to perform, resulting in increased disfluency. For those children who are really struggling with stuttering, parents’ focus on verbal performance may create in the children feelings of shame at their verbal ineptitude.
It is useful to compare a child’s language (syntax) scores with his vocabulary scores. Researchers (e.g., Anderson & Conture, 2000; Conture, 2001) have shown that many children who stutter have a disparity between syntax and vocabulary scores that is greater than that for matched typically developing children. Interestingly, it has also been shown that children who have lower language abilities at the beginning of treatment show greater long-term decrease in stuttering as a result of treatment (Richels & Conture, 2010).
As I review my observations of a child’s speech and language, I consider not only the possibility that a child’s language is advanced relative to his speech motor abilities, but also the possibility that his motor abilities are markedly delayed. A few children have motor problems that impair their coordination of respiration, phonation, and articulation with language production. Many are aware that speech is difficult for them and have already felt frustration and shame, not just about stuttering, but about the way they speak and how they perform other fine motor tasks. Therefore, to help these children improve their feelings about themselves as talkers, the parents and I work on their speech motor skills. These children seem to benefit especially from models of slow speech as well as activities that teach them to speak more slowly.
In addition to exploring the possibility of language and articulation difficulties, I also assess a child’s voice. A hoarse voice may be especially significant in a preschool child who stutters because it may be a sign that the child has increased tension in his laryngeal muscles, perhaps in an effort to cope with stuttering. I look closely at how the child is handling his blocks and listen for signs of excess laryngeal tension, such as pitch rises, increases in loudness, and hard glottal attacks. Because many of the techniques I use in treatment of stuttering result in a more relaxed style of speaking, I usually don’t treat voice separately from stuttering. However, if a child has voice problems other than hoarseness, or if hoarseness does not diminish with stuttering therapy, I refer the child to an otolaryngologist and follow treatment approaches such as those suggested by Boone, McFarlane, Von Berg, and Zraick (2009).
Other Factors
In Chapters 4 and 5, I described a number of possible developmental influences on stuttering. In this section, I review them briefly so that they may be recognized if they are important in a particular preschool child’s stuttering. Because much of this information can be obtained from a parent interview, you may wish to consult
Chapters 4 and 5 for further details on developmental influences before conducting a parent interview.
Physical Development
I like to ascertain whether a child has age-appropriate gross motor skills and whether his oral motor development is typical. Figure 4.2 in Chapter 4 presents information about motor development. Most children learn to walk at about age 1 but usually do not learn to walk and talk at the same time. If a child I am evaluating was delayed in walking but average or advanced in talking, I may explore the possibility that the onset and evolution of his stuttering was associated with his delayed motor development.
Cognitive Development
When I consider a child’s cognitive development, I want to learn whether there is cognitive delay, which can be associated with increased disfluency. I also want to know if a child may be going through a period of rapid cognitive growth that might, hypothetically, take a temporary toll on fluency.
Social-Emotional Development
As a child grows, various tensions develop between him, his parents, and his siblings. Between ages 2 and 5, many children may display negativity in ways that are felt throughout the family. When I ask a child’s parents about conditions surrounding the onset or worsening of a child’s stuttering, I explore social-emotional factors as well as environmental and developmental factors.
In Chapter 4, I described various life events that may affect a child’s stuttering. In the parent interview, I examine life events surrounding the onset of stuttering to see if upsetting events or ongoing situations may be linked to the child’s stuttering. Some events, like the birth of a sibling, may be happy ones, but they can create disturbances in the psychological balance of a family.
Speech and Language Environment
I have referred to the child’s communication environment before, but here I am more explicit. Many children have their hands full trying to compete verbally with fasttalking, articulate adults. Children who stutter may find this particularly hard. If the family has sent, as requested, a video or even audio recording of the parents and child talking, I observe the parent-child interactions carefully for indications of a complicated verbal environment, such as rapid speech models without pauses that may be like rough water to a new swimmer.
Now let us turn to the task of pulling together the information gathered in an assessment and making a diagnosis of a young client’s problems. One of the clinical questions that must be answered is whether the child is truly stuttering. Once you have (tentatively) answered that, you can answer the other clinical questions and describe the child’s stuttering, his reactions to it, and an appropriate treatment choice.
Determining Developmental and Treatment Level
In determining an appropriate treatment for the child, I begin by trying to figure out if a preschool child is normally disfluent and if not, his level of stuttering: borderline or beginning stuttering. In the following paragraphs, I briefly review these levels, which were described in detail in Chapter 7.
Typical Disfluency
All of the following characteristics must be met for a child to be considered normally or typically disfluent. The child has fewer than 10 disfluencies per 100 words; these disfluencies consist mostly of multisyllable word and phrase repetitions, revisions, and interjections. When disfluencies are repetitions, they will have two or fewer repeated units per repetition that are slow and regular in tempo. The ratio of stuttering-like disfluencies to total disfluencies will be less than 50 percent. All disfluencies will be relatively relaxed, and the child will seem to be hardly aware of them and certainly will not be upset when he is aware.
A child may be considered to have borderline or beginning stuttering if he has any of the characteristics described in the following paragraphs. Place him at the level– borderline or beginning–that includes the child’s most salient characteristics.
Borderline Stuttering
The child I place in this category has more than 10 disfluencies per 100 words, but they are loose and relaxed. They may be part-word repetitions and single-syllable word repetitions, as well as prolongations, and the repetitions may have more than two repeated units per instance. Stuttering-like disfluencies will be above 50 percent (Yairi, 1997a), and the disfluencies may cluster on adjacent sounds (LaSalle & Conture, 1995).
Beginning Stuttering
Beginning stuttering usually occurs in children between 3.5 years old and 6. The key features at this level are the presence of tension and hurry in the child’s stuttering. Disfluencies may have some of these characteristics: rapid, abrupt repetitions; pitch rises during repetitions and prolongations; difficulty starting airflow or phonation; and signs of facial tension. Just the occasional appearance of these signs would make me believe the child is a beginning stutterer. A beginning stutterer also shows that he is aware of his stuttering (in some, this may be subtle) and may be frustrated by it. He may use a variety of escape behaviors, such as head nods or eye blinks in terminating blocks. Occasional avoidance may occur. For example, a child who has developed language to the point of using “I” instead of “me” but begins to stutter on “I” at the beginnings of sentences may begin substituting “Me” for “I” to avoid the frustration of stuttering on “I.”
Some children are relatively advanced for beginning stutterers. These are children who avoid words and situations, and their behavior and demeanor clearly suggest some fear and shame about stuttering. For example, they may use a variety of starters to begin sentences and look away or appear embarrassed when they stutter.
Although I use information from all sources to determine a child’s developmental and treatment level, I have found that my own observations of parent-child and clinician-child interactions are most useful in making this (tentative) decision. Parents are helpful in describing long-term changes in their child’s stuttering, but they frequently miss avoidance behaviors, such as starters, circumlocutions, and postponements, which are critical indicators of this more advanced level of stuttering. Parents’ reports do provide, however, as much information about a child’s feelings and attitudes as I usually gather in observing interactions in the clinic. Thus, parent reports plus my own observations provide valuable, complementary data. A vital adjunct to direct observations are video recordings of parent-child and clinician-child interactions. I sometimes revise my initial placement of a child in a developmental/treatment level after viewing video of the interactions I have already directly observed.
Risk Factors for Persistent Stuttering
Risk factors are those elements within a child or in his environment that make it more or less likely that he will persist in his stuttering (Guitar & Guitar, 2003) or take longer in treatment. Table 9.1 describes several of these factors. For some risk factors for persistent stuttering, there are data which support this connection; for others, only clinician observations provide support. I described much of the evidence for these factors in the chapters on Constitutional Factors in Stuttering (Chapters 2 and 3) and on Developmental, Environmental, and Learning Factors (Chapters 4 and 5). You’ll be able to get the information you need to determine a child’s risk of persistent stuttering from the case history form, questionnaires, parent interviews, and observations of the child.
Table 9.1 Risk Factors for Persistent Stuttering or Extended Treatmen

Drawing the Information Together
After I have completed the assessment tasks, I consolidate the information I have gathered, develop a tentative diagnosis, and meet with the family in a closing interview to discuss my findings and their desires and expectations. If I had been able to get a recording from the parents beforehand, I would have carefully analyzed it before the assessment and would present the results of my analysis in the closing interview. The same is obviously true for the other information I obtained before the assessment, such as the Behavioral Style Questionnaire (McDevitt & Carey, 1995) that was described in Chapter 8. Some data, like my analysis of the recordings of parentchild and clinician-child interactions, may not be completed by the time of the evaluation, but results will be included in my assessment report. Thus, for the closing interview, I rely on a combination of previously acquired quantitative data, some of the quantitative data I acquired as I talked with the child, such as frequency of stuttering, and the qualitative data gleaned from my observations and interview questions.
Prior to the closing interview, I spend a few minutes studying my findings or discussing them with students or colleagues if I have been working with a team. I try to make sure that I have obtained enough information to answer two key questions: What is this child’s developmental/treatment level of stuttering, and what is the appropriate treatment? Working with the family, we can decide where to go from here. Last year, I began a practice of writing up a brief, one-page summary of our findings and recommendations prior to the closing interview. I share a copy of this recommendation sheet with the family so that we will have a common reference as we discuss the evaluation and where to go from here. They can share this with other family members and if appropriate, the family can begin to make some changes immediately, without having to wait for the formal written report, which may take several days to prepare.
Closing Interview: Recommendations and Follow-Up
Before I begin, I remind myself to take the necessary time to listen to the family’s questions that may arise at any time during the closing interview. When I begin, I always make some positive comments about the child and the family, then describe to the family characteristics of the child’s stuttering that I observed in parent-child and clinician-child interactions and the preassessment recording, if I obtained one. I stay away from jargon and strive to be as clear and straightforward as possible as I briefly describe the child’s behaviors, review the more important information that the parents provided in the case history and our interview, and estimate the seriousness of the child’s problem. If stuttering is a serious concern, I say so, and if the parents have expressed feelings of guilt about their child’s stuttering, I again reassure them that they are not to blame but that they will be crucial in helping to resolve it. Next, after answering questions, I describe appropriate treatment approaches, such as environmental changes, indirect treatment, and direct treatment, which will differ depending on the developmental/treatment level of the child’s problem.
Recommendations for Children with Typical Disfluency
If I believe that a child’s speech is typically disfluent, I deal with the family’s concerns rather than the child’s disfluencies. Most families benefit from knowing how I reached my tentative conclusion, so I provide them with information about normal disfluency, such as the following: “During their preschool years, many normal children pass through periods of disfluency. Interjections, revisions, pauses, repetitions, and prolongations are common during these periods, but they usually occur in fewer than 10 of every 100 words. Interjections and revisions are more common than part-word repetitions, and part-word repetitions usually have only one or two repeated units per disfluency. Children who are normally disfluent are largely unaware of their disfluencies, do not react negatively to them, and gradually ‘outgrow’ them.”
In most cases, I use analogies to help the family understand their child’s disfluent speech. For example, I may point out that learning to speak is like learning many other skills, such as riding a bike or learning to skate, and that a learner falls down a lot in the early stages. I look for analogies that will fit the family’s experiences to help them understand why their child is disfluent and how valuable an accepting environment can be for a child’s self-esteem. Parents who are concerned about their child’s typical disfluencies usually feel reassured when they find out that this is not uncommon. In those rare cases when parents are still not convinced that their child’s speech disfluencies are not stuttering, I teach them how to slow their speaking rates, pause frequently, simplify their language, and relieve other pressures that we mutually agree to change. Then, I set up another appointment to discuss their progress and the child’s speech. If parents really are concerned and seem likely to continue worrying and perhaps correcting their child’s speech, a few sessions focused on the normalcy of their child’s speech and the changes they have made in the family’s environment may be an ounce of prevention.
I keep the door wide open for all parents of normally disfluent children. I reassure them that I am available to talk with them if they become concerned again and will be ready to work with them if their child does begin to stutter.
Recommendations for Children with Borderline or Beginning Stuttering
For those preschool children evaluated less than 12 months after the onset of stuttering, there are guidelines to help decide which children should begin treatment and which can be followed for a period of time without treatment. First, children whose stuttering-like disfluencies (part-word and single-syllable word repetitions, prolongations, or blocks) steadily decrease during the first 12 months after onset are more likely to recover without formal treatment. There are more indicators of the likelihood of recovery without treatment, listed in Table 9.2.
Table 9.2 Factors that May Be Associated with Increased Likelihood of Recovery from Stuttering without Treatment*

I believe that any preschool child who has borderline or beginning stuttering should be treated or followed carefully for several months. I stay in telephone or e-mail contact with families of children who are close to onset whose stuttering is diminishing, who have other indicators suggesting recovery without treatment is likely, and whose families are not overly concerned. However, if families are highly concerned or the child’s stuttering is not decreasing and there are few indicators of recovery, I begin treatment as soon as possible.
My closing interview with these parents is usually the first of many sessions we will spend together. Consequently, I don’t need to accomplish everything in this meeting. Because treatment of any preschool child who stutters is often focused on the home environment, we frequently begin our discussion with things the parents can do at home. The chapters on treating preschool children (Chapters 11 and 12) contain more extensive discussions and guidelines for involving the family in treatment, but I will make some initial suggestions here.
In my experience, parents who are active in the ongoing assessment process from the beginning feel more hopeful, less guilty, and more motivated to be involved in treatment (see also Zebrowski & Kelly, 2002). Therefore, in the closing interview, I ask parents of preschool children who will soon start treatment to begin observing and recording the day-to-day variations in their child’s fluency. Having them assess fluency in the home environment also gives me a more valid indication of changes in stuttering than if assessments are done only in the clinic. I teach parents to use the Severity Rating Scale (Fig. 8.4; Onslow, Andrews, & Costa, 1990; Onslow, Packman, & Harrison, 2003), which is a form on which they record, at the end of each day, a number from 1 (no stuttering) to 10 (extremely severe stuttering), which is their estimate of the severity of their child’s stuttering. Parents begin by rating the severity of the child’s stuttering during the parent-child interaction sample just recorded. The clinician also rates the severity of this sample. If the parents’ and clinician’s ratings differ by more than one point, the parents and clinician discuss the ratings and watch the recording of the interaction to help them come to a consensus. If more than one parent or another family member will be using the Severity Rating Scale, each person should be trained until his or her ratings are within one point of the clinician’s rating of each sample. More discussion of using this rating scale is provided in the chapters on borderline and beginning stuttering. In addition to rating the child’s severity every day, parents can record comments and questions they would like to discuss when we meet at the next session.
If a child with borderline stuttering is being followed but not formally treated, severity ratings are an important part of the monitoring process. Clinicians can obtain information about the child’s stuttering via phone calls or e-mail on a regular basis, and parents can report their severity ratings for each day as well as discuss issues of concern and ask questions. In addition to monitoring the severity of a child’s stuttering, it is often helpful to brainstorm with the parents about ways in which the environment might be made as facilitating as possible for their child’s speech. I shall discuss this in more detail in the following paragraphs.
For the younger preschool child with borderline stuttering who is being treated (i.e., a child whose parents are very concerned or who has multiple risk factors for persistent stuttering), the closing interview is a time when further appointments may be set up and changes in the family environment can be initiated. Such changes will be determined by the clinician’s observations of parent-child interactions, the parent interview, and ideas that parents may have about what they would like to change. In my experience, one of the most powerful ways that parents can facilitate fluency is to set aside 10 to 15 minutes each day, preferably in the morning, for child-directed interactions. This is a one-on-one interaction without other children interrupting. In two-parent homes, parents may need to alternate which one does the one-on-one activity so that the other parent can watch the other children. Or a parent may conduct the session when the siblings are at school or napping. During these interactions, the parent primarily listens to the child and plays whatever games the child chooses. When the parent speaks, he or she should use a slow rate with frequent pauses, somewhat like television’s Mr. Rogers. I have found this works best if the clinician models this interaction style and then watches the parent carry it out. More information about changing the family environment is given in the chapter on treatment of the young preschool child with borderline stuttering (Chapter 11).
I also give parents of these children reading material or a video recording to help them better understand stuttering and what they can do to help their child. The book, Stuttering and Your Child–Questions and Answers (Conture, 2002), gives many good suggestions and is available from the Stuttering Foundation of America ( for very little money. The video, Stuttering and the Preschool Child-Help for Families (Guitar & Guitar, 2003, SFA publication no. 70), is also available from the foundation, both through the online store and as a streaming video on Another video, Preventing Stuttering in the
Preschool Child: A Video Program for Parents (Skinner & McKeehan, 1996; Communication Skill Builders) is highly instructive.
For older preschool children with beginning-level stuttering, I begin treatment as soon as possible. For stuttering at this level, I use a direct approach. Options for therapy should be described to the parents, and with the clinician’s guidance, they should make an informed choice. In some cases, the family will be able to make their decision immediately. In other cases, they may need time to consider the possibilities. If they choose to begin treatment soon, one or more family members should be trained immediately in recording daily severity ratings, and the next session should be scheduled. They should be asked to bring in their severity ratings if the session could be scheduled within a week or two. If it has to be delayed, the clinician should be in contact with the family through e-mail or telephone each week to discuss their severity ratings until formal treatment can begin. Once that begins, parents can bring in their severity ratings each week to discuss them with the clinician.
If treatment cannot begin for several weeks, I ask a parent or another family member to conduct one-on-one interactions with the child as was just described for a child with borderline stuttering. If the family is able to begin treatment immediately, the clinician should start the parents on appropriate activities. Clinicians who carry out therapy themselves with the child will probably have a parent watch the first few sessions before beginning direct activities at home. My own preference is to use the Lidcombe Program (Harrison & Onslow, 2010), which is a parent-delivered treatment. Consequently, if therapy can begin immediately, I describe the first phase of this treatment program to the parents, which is a daily parent-child session conducted at home. Parents engage the child in an activity at an appropriate linguistic level to elicit fluent speech and reinforce fluent utterances. After explaining this to the parent, I model this type of interaction and then observe the parent as he or she tries it. The Lidcombe Program and other direct and indirect approaches are described in Chapter 12.
The closing interview should end when the family seems to have a good understanding of the clinician’s findings, and they and the clinician agree what the next steps should be. Because a family may come up with new questions and concerns in the days following the evaluation, it is important to conclude the interview with information about how they can contact the clinician.
Clinical Questions
As with preschool-age children, it is important to begin an evaluation with certain questions in mind. What is this child’s frequency of stuttering? What types of disfluencies does he display, and what is the percentage of SLDs? What is the child’s severity? What is his speech rate? With rare exceptions, the question of whether the youngster is normally disfluent or stuttering is not an issue. By age 6, most children who stutter do so in ways that are quite different from the normal disfluencies typical for their age. Another question is what emotions and attitudes does the child have about stuttering and about speaking? School children with notable fear and avoidance may need special attention to these feelings and behaviors. Information about risk factors (e.g., gender, family history) are important but not as critical as they are for a preschool child. By the time a child is in school, natural recovery is less likely than in the preschool years; thus, an absence of risk factors doesn’t warrant withholding or delaying treatment.
Information from the child’s teachers, the clinician’s observations of his speech in class and in the treatment room, and information from his family are all required to assign the child a developmental/treatment level. Questions about treatment of children in the public schools can be answered only in the context of federal and state laws, which are considered in the next section. With any school-age child, it is vital to determine the child’s school performance and how stuttering interferes with it.
Public School Considerations
The Individuals with Disabilities Education Act (IDEA ’97 and changes made in 2004) and individual state laws mandate the procedures that public school clinicians must use for gathering information about a child’s disability and deciding on treatment. In most states, a “prereferral prevention/intervention” process is used when a teacher encounters a child who stutters in the classroom (Moore & Montgomery, 2007). Speech-language pathologists are usually members of a team that consults with the teacher and parents to determine if a child’s difficulty can be resolved by making changes in the educational setting. An example of such modifications might be discussions between the child and teacher about how the teacher can facilitate the child’s class participation. If stuttering continues to be a problem in the classroom after the modifications have been in place for a designated time period, the teacher or parent can refer the child for further evaluation. The next step, evaluation by a multidisciplinary team, is usually taken in response to a referral or as the result of a clinician’s identification of a child through screening. As part of this evaluation, the clinician discreetly observes the child in the classroom and confirms (or disconfirms) that the child is stuttering. The clinician then discusses the child’s problem with the teacher and the school’s special education administrator. Next, the clinician, teacher, or administrator contacts the child’s parents to ask permission to do a formal evaluation of the child. If permission is given, the clinician gathers information on as many dimensions of the child’s stuttering as possible. Typically, this will include the frequency, severity, and types of stuttering observed in two or more situations, the child’s feelings and attitudes about stuttering and speaking, concomitant speech or language problems, and overall communicative performance. The clinician uses standardized tests such as the SSI-4, observations, and interviews with the child and his family as well as with his teachers and others at school who know him. After this information is gathered, a team composed of the clinician, teacher, special education administrator, and the parents meet to decide two issues. The first is whether the child’s stuttering problems meet the state’s criteria for eligibility, and the second is whether the child’s stuttering adversely affects his educational performance. These two issues are discussed in detail later in this chapter, after the sections on the parent, teacher, and child interview.
Initial Contact with Parents
Whether contact is made because the child has been referred to the school clinician or because the parents have made an appointment at a private clinic, the clinician’s most important task is to listen and try to understand the parents’ point of view. If the school clinician is telephoning the parents for permission to evaluate their child, she should describe the process by which the child was identified and convey her and the school’s desire to help the child achieve his potential as an effective communicator. It will be helpful to briefly describe the disfluencies that identified the child as stuttering and to find out if the parents have also noticed them. The clinician should calmly convey her interest in the child and his stuttering in an accepting tone of voice, particularly because parents may fear that they are being blamed for the child’s stuttering. It may help also to comment that current views suggest that stuttering may be the result of how the child’s brain is organized, although its exact cause is unknown. The evaluation process should be described and permission sought. If the parents agree to an evaluation, this is a good time to ask them to fill out a case history form and if possible, send a video recording of the child’s speech. It may be beneficial for the clinician to talk to the child as well and ask his permission to have his parents video record his speech at home. In some cases, the home video is easier to obtain once the clinician has gotten to know the child and has conveyed her acceptance and interest in the child’s stuttering.
In cases in which parents have made an appointment for an evaluation at a clinic, the clinician should call the parents and let them know what will take place in the evaluation, get some preliminary information about the child and his stuttering, let them know they will receive a case history to complete and return, and request a video from home prior to the evaluation. As with the school clinician’s telephone call, the parents’ point of view about stuttering must be understood. Even though they will have a chance to talk over their concerns in person, they may also want to talk and ask questions in this preliminary telephone call.
Case History Form
The form used for this age group is the same one used for preschool children (Fig. 9.3). Some of the questions about speech and language development may be difficult for parents to recall. This is not critical for evaluating a school-age child, but it is important to probe for other speech and language problems that may be contributing factors in the school-age child’s stuttering. An important section on this form deals with how the problem has changed since it was first noticed, what has been done about it, and how others have reacted to it. In addition, the section on educational history lets us know if the child is having problems in school.
Audio/Video Recording
Obtaining a recording (preferably a video recording) of the child speaking at home or elsewhere will help clinicians prepare for the evaluation because they can get a preview of the child’s pattern of stuttering, analyze the sample ahead of time, and plan the assessment more carefully. For example, if a sample from home has little or no stuttering, the clinician may want to obtain another sample in a more difficult speaking situation. Up to a point, more varied samples of a child’s speech lead to a more valid assessment. If a pre-evaluation sample has lots of avoidance behaviors on it, clinicians can prepare questions to ask the child about what he does when he expects to stutter.
Parent Interview
This description of the parent interview assumes that the parents have brought their child to a clinic for the evaluation. When the evaluation is school-based, the clinician can get much of this information by telephone and follow-up with a face-to-face meeting at school.
Begin a clinic-based interview by sharing some positive observations about the child and his family and then describe the course of the evaluation. Before obtaining more background information to fill in gaps left by the case history, ask parents an open-ended question, such as requesting them to describe their concerns about their child’s speech. Only after the parents or caregivers have had a chance to express their worries and observations, do I ask follow-up questions. As I do with parents of preschool children, I explore the onset and development of the child’s stuttering, his reactions to it, family members’ reactions, and any gaps in the case history. I also ask parents of a school-age child about his school experiences. Does he like school? Does his speech seem to bother him there? Do you think he participates less in school because of his stuttering? Is he teased or bullied about his stuttering? Do you think he stutters more at school than at home? Has he gotten therapy at school? Has that helped? Has he liked it?
As I ask parents questions about the child’s stuttering at home and school, I listen for responses that may help me understand why the child’s stuttering has persisted into elementary school. Here are some of the questions I think about as I try to assimilate the information I am getting from the parent: Is the child sensitive about his stuttering? Are the family and child comfortable talking about stuttering? Is the family supportive of the child and his ways of coping with his stuttering? Is the family motivated to participate in therapy?
I also keep in mind that the parents are probably doing their best, given the expectations with which they grew up. One of the most important things clinicians can do in parent interviews is to convey an acceptance of the parents as they are and to point out the helpful things they have done for their child.
Teacher Interview
The more assistance we can get from a child’s teachers, the more we can help the child. We need to approach teachers respecting their heavy responsibilities and their concern for their students, including the one with whom we are working. But we also should anticipate that they may neither understand nor know what we do to help a child who stutters. As I conduct a teacher interview, for example, I try to sense what they would like to know about stuttering and my treatment approach. The following questions serve as guidelines for the types of things I want to find out.
1. Does the child talk in class? Does he stutter? What is his stuttering like? How does he seem to feel about his stuttering and about himself as a communicator?
Here, I am trying to determine how much the child stutters in class and whether his stuttering keeps him from talking as much as he might otherwise if he did not stutter. I may also get a flavor of how the teacher feels about the child, his communication abilities, and his stuttering.
2. Does stuttering interfere with the child’s performance in school?
This question is obviously related to the previous one about the child’s stuttering in class. But it also may give us some information about how much the child may avoid speaking, especially volunteering in class. I ask about disparities between his oral and written performance; a large disparity may indicate that he declines to talk or says “I don’t know” even when he knows the answers.
3. Do other children tease him about stuttering?
Most school-age children who stutter are teased, and I want to get more information about how much he is teased and how it affects him. I also want to find out about any school policies that relate to bullying since teasing may be just the tip of the iceberg, and many schools are developing strategies for addressing both problems.
4. How does the teacher feel about stuttering, and how does he or she react to it? I am often able to get this information indirectly, from what he or she said before, but if not, I ask directly. Teachers are also likely to ask how they should respond to a child’s stuttering, which is an important issue because a teacher’s response often influences how the class responds. This and other issues related to the child’s speech in the classroom are discussed in the chapter on intermediate stuttering (Chapter 13).
Classroom Observation
In addition to the information obtained from teacher and parent interviews, direct classroom observation can help clinicians understand the severity of a child’s stuttering and the degree to which it interferes with his academic adjustment. If a child is to receive services in the school, the clinician must establish that the child’s stuttering is interfering with his education. One way to verify this is by firsthand observation of a child in the classroom.
You should arrange with the teacher to come to the classroom at a time when the child will have opportunities to participate in class and observe the class as unobtrusively as you can. By observing the class when many students are
participating, not just when the child you are evaluating is talking, you will not call as much attention to him. Most school-age children want to be like their peers and dread being singled out. Notice whether the child participates in class discussion. If called on, does he speak in a straightforward manner or does he hesitate or deploy any postponement devices, such as repeating “uh” several times? Does he answer “I don’t know?” This is a reply many children who stutter (including me) use to avoid speaking and thus risk stuttering–even when they know the answer. If the child does talk in class and happens to stutter, notice how other children react to his stuttering. Do they giggle and look at each other and make comments under their breaths, or do they seem normally attentive? Because the classroom is the arena where children learn, socialize, and develop communication skills, it should be a target of assessment and treatment.
Child/Student Interview
After I obtain parents’ consent to evaluate a child, I arrange for the child to come to the treatment room. Here, I set about to make him feel that my room is an accepting environment where he can have fun and also discover how he can make his speech much easier. School-age clients sometimes tell me that it helps just to have someone to talk to about stuttering and other things that are bothering them. This can occur only after a trusting relationship is established, and the initial interview is the beginning of building that relationship.
In our first encounter, it is important for a child to feel that I am genuinely interested in him as well as his stuttering. I usually begin by asking what he likes to do, who his friends are, and who is in his family. Then I tell him a little about myself and how I work with other kids who sometimes get stuck on words. As the child talks, I note whether he stutters or not and how he stutters. When a child’s body language and behavior tell me he’s comfortable in the session, I talk to him about his speech. The following questions are not asked one right after another but over a session or two. Often, it is more effective to make the question a comment, such as phrasing the first question below as: “Sometimes kids have trouble getting words out. Their words just seem to stick a little bit.” Then, leave some silence to see if the youngster will talk about his own speech.
1. Do you ever think that you have any trouble talking?
I rarely see school-age stutterers who are unaware of their difficulty. However, if achild regards his problem as minor or seems genuinely unaware of the problem, I avoid giving it undue emphasis or creating an unfavorable attitude about it. Thus, my first talk with a child is usually low-key, and if he truly doesn’t seem to be bothered by his stuttering, although his parents and teachers are, I respect his perception and try to treat it as a relatively minor problem, but I remain aware that the child may be bothered by his stuttering much more than he wishes to let on at first.
2. What happens when you get stuck on a word? When does it happen? Is itdifferent at different times?
I am looking for several things here. One is to learn the words the child uses todescribe his stuttering so that I can use them when talking with him about it. I also want to find out if the child is unaware of some of his stuttering behaviors, if they seem to be too painful for him to face, or if he just doesn’t like talking about them. Even more important, these questions let a child know that the clinician really wants to understand his problem.
3. Have you learned to use any helpers or “tricks” to get words out? Do you sometimes avoid certain words?
With this question, I can convey that I understand what some people do when they stutter. I can also let a child know that I am nonjudgmental about the “tricks” he uses by conveying my acceptance and interest in his descriptions. In addition, I am also exploring which level the child’s stuttering has reached by determining if he is using escape and avoidance behaviors.
4. Are certain speaking situations more difficult?
This is another question that helps me understand what a child is experiencing, while conveying my understanding.
5. Most kids who stutter get teased or picked on about their speech. Do you everget teased about your stuttering? What do you do when that happens? How does it affect you?
Many children who stutter are teased about their speech but are not willing to talk about it straightaway with someone they don’t know well. So this question is a “feeler,” and if the child denies being teased, the clinician should not dwell on it now.
6. How do you feel about your speech?
To help a child express feelings about stuttering, I can suggest some possibilities by asking, “Does it make you mad sometimes?” or “Do you wish you didn’t get stuck?” Don’t be surprised, however, if a child says it doesn’t bother him because his feelings may have been rejected, perhaps unintentionally, by adults. Adults may say, for instance, “You shouldn’t feel that way,” or “Why do you let it bother you?” An effective clinician will show the child that whatever feelings he has are OK and that the clinician is really trying to understand. Real discussions of feelings probably won’t begin until a child has learned to trust the clinician deeply. However, in this first interview, I may be able to infer what some of the child’s feelings are and from that, understand how far his stuttering has advanced. Another avenue to elicit feelings is through drawing pictures. For some children, drawing makes it easier to talk about feelings. The child doesn’t have to look directly at the clinician, and his self-consciousness may be decreased by his focus on drawing. I usually suggest to a child that both of us draw whatever we would like, and as we are drawing, I talk about feelings. If this goes well, I bridge the gap between the drawing and talking by suggesting that the child might want to draw a picture of what stuttering is like or what he feels like when he stutters. I have found that this technique can make extensive discussion of feelings much easier for some children. In some cases, children have used their drawings when they talked about stuttering with their class once therapy has helped them feel more comfortable with themselves and their speech.
Some of the activities in the workbook, The School-Age Child Who Stutters: Working Effectively with Attitudes and Emotions (Chmela & Reardon, 2001), are helpful in exploring a child’s emotions in both the evaluation and treatment. I shall discuss some of these a little later.
7. How do your parents feel about your speech? Do they ever say anything orgive you advice?
This helps me determine what sorts of experiences the child may have been going through at home. One parent may be much less accepting of a child’s stuttering than the other. Whatever I find out may help me enlist the parents’ participation in treatment.
8. Can you think of anything else important for me to know about you or aboutthe trouble you sometimes have when you talk?
This lets the child know that I am interested in him and that his ideas are important to me.
Speech Samples
With a school-age child, I video record him talking for 10 minutes about school and other activities in the therapy room. I prefer not to turn on the recorder the moment the child walks into the room. Instead, after talking for a few minutes, I ask the child if he would mind my recording our conversation as we talk. If it’s OK with the child, I record a sample that optimally includes 300 to 400 syllables of his speech. For those few children who are reluctant at first, I explain that I need a recording of their speech to understand their stuttering better. In rare cases, I might need to postpone recording until the child is more comfortable with me. After recording the speech sample, I ask the child to read approximately 200 syllables of age-appropriate material. I often use the SSI-4 examiner’s manual, which has 200-syllable reading passages at the third, fifth, and seventh grade levels.
If possible, I also obtain a video recording from home for a second sample. In some instances, you may not be able to get a second recorded sample from home or elsewhere. Even if you are not able to record a second sample, write down your impressions of the child’s stuttering, including the amount of stuttering and the core and secondary behaviors you observed.
Pattern of Disfluencies
A school-age child is likely to show beginning or intermediate stuttering, so you want to know as much as possible about the amount of tension in the child’s stuttering, the escape behaviors he uses, and the extent to which he avoids words and situations. You can obtain this information directly by observation or indirectly through parent and teacher interviews. As with adults or adolescents who stutter, I use this information not only to decide at which developmental/treatment level to place the child but also, when appropriate, to plan the process of unlearning conditioned responses, which, once created, now maintain the child’s pattern of stuttering.
Stuttering Severity Instrument (SSI-4)
Samples of conversational speech and reading are needed to calculate scores on the SSI-4. Administration and scoring of the SSI-4 were described in Chapter 8.
Speech Rate
The samples you collect for rating severity with the SSI-4 can also be used to assess the child’s speaking rate. The purpose of assessing speech rate is to get some idea of how much the child’s stuttering interferes with the rate of speech he normally uses. As I help the child manage his stuttering, I expect a steady increase in his speech rate toward normal levels.
Normal speech rates for school children in Vermont between the ages of 6 and 12 years measured in syllables per minute are given in Table 8.5. These rates were obtained from children’s conversations with a clinician about Christmas, hobbies, school, and home activities. They were calculated by including normal pauses in their conversation but excluding pauses longer than two seconds for thought. It is reasonable to expect that children’s speech rates in other states will be similar.
Trial Therapy
Trial therapy with a school-age child will help me understand what approach might work with this particular child as well as what may be difficult for him. If the child is able to make significant, though temporary changes in his stuttering in this brief treatment, he will gain hope and motivation for our work together.
I usually begin by asking the child to identify moments of voluntary stuttering in my own speech. I explain that I will be putting stutters in my speech and want to see if he can catch me. It’s more fun if I can use a small reward for his successes. I then tell him about my favorite recent movie or television show (something he can later talk about himself) and put in a variety of stutters in my speech. As I talk and encourage him to catch my stutters, I let him know how good he is when he catches one of my stutters without my help and hand him a piece of candy or another reward. After a few successful catches, we switch roles and have him put in some stutters–pretend or real –as he talks about a movie, TV show, or any other handy topic, and I try to “catch him.” Each time, I make a positive comment about his stutter and give him a reward. Rewarding stutters of school-age children causes no harm; in fact, it reduces negative emotion, which is a very positive step in treatment.
A next step is to see if he can hold on to a stutter. As always, I first demonstrate what I’m asking the child to do. I have the child signal me by pointing to me when he wants me to stutter and make me hold on to the stutter for several seconds by continuing to keep his finger pointing at me. Then, we reverse roles and I coach him to hold on to either voluntary or real stutters. In so doing, I also coach him, while he’s holding on to the stutter, to let it go slowly and loosely when he’s ready to move on. Typically, my models, my enthusiasm, and the reinforcements I use enable most children to be able to carry out these activities. Note that if the child cannot do these activities, it suggests a higher level of fear or an inability to focus on the task. These possibilities usually mean that a child needs a slower approach, and I will consider teaching him fluency skills before attempting stuttering modification.
Trial therapy using fluency skills simply involves teaching the child one or two of the fluency skills described in Chapter 13 on treatment of stuttering in school-age children. I use a word list and give the child an example by producing each word myself before he tries it, using a slow rate and gentle onset of voicing. The severity of the child’s stuttering will determine how slowly I begin the word; my aim is to use modeling of slow rate and easy onset to produce fluency in the child. Once he can say words after me in that slow fashion, I then ask him to say each of several words again, but without my model. If he can do this, I create sentences beginning with those words said slowly and with an easy onset (but with the remainder of the sentence produced at a near-normal rate) and again assess whether he can repeat them fluently with my model and then without.
These exercises help me determine how well the child can make changes in his speech and his stutters. By using a large amount of modeling and appropriate reinforcement, I can often take the child quite far along in the time I have.
Feelings and Attitudes
A fair assessor of a child’s feelings and attitudes is the clinician’s judgment, which usually improves as she gets to know the child better. Nevertheless, the clinician should be able to get a pretty good indication of a child’s feelings and attitudes from the first interview. Some indications will emerge from your discussion with the child about his feelings, and some will emerge from observations of his behavior. Watch how the child responds when asked about his stuttering, and note how much he avoids stuttering. When the child does stutter, observe how calm he is and how consistent his eye contact is.
After the clinician has gotten to know the child a bit, she may want to administer a paper and pencil assessment of attitude. Figure 8.6 in the previous chapter depicts the A-19 scale (Guitar & Grims, 1977), a measure developed to assess children’s communication attitudes. This scale consists of questions that were found to distinguish children who stutter from those who do not. Hence, if treatment is effective, a child’s attitude about communication may change, although this has not been established by research.
In addition to the A-19, the Communication Attitude Test (Fig. 8.7), which was developed by Brutten, has been tested on nonstuttering children (Brutten & Dunham, 1989) and shown to differentiate them from stuttering children
(De Nil & Brutten, 1991). It has also been found to have good test-retest reliability for this purpose (Vanryckeghem & Brutten, 1993).
Many informal methods of assessing feelings and attitudes are given in the workbook, The School-Age Child Who Stutters: Working Effectively with Attitudes and Emotions (Chmela & Reardon, 2001). These include such activities as a “Worry Ladder,” in which a child lists his worries in a hierarchy, and “Hands Down,” which elicits things the child likes and does not like about himself. Although the reliability and validity of these tools have not been determined, they provide useful starting points for communication about feelings and attitudes.
With some children, both formal and informal methods of assessing feelings will be productive during the evaluation. But others will hold back until they have developed a trusting relationship with the clinician. Thus, clinicians should be mindful that information about a child’s feelings and attitudes obtained in a first meeting may not be complete or accurate.
Other Speech and Language Disorders
In my discussion of the preschool child, I described the importance of screening language, articulation, and voice. The same abilities should be screened in the schoolage child. Conture (2001) suggests using the “Sounds in Words” subtest of the Goldman-Fristoe Test of Articulation 2 (Goldman & Fristoe, 2000) for articulation, the Clinical Evaluation of Language Fundamentals-4 Screening Test (Semel, Wiig, & Secord, 2004) for language, and the Peabody Picture Vocabulary Test 4 (Dunn & Dunn, 2007) and Expressive Vocabulary Test 2 (Williams, 2007) for vocabulary. The child may have a previously diagnosed language or articulation (or to a lesser extent, voice) problem and may be in therapy. If so, the clinician should seek out details of any current or previous therapy. If a child is receiving or has received articulation or language therapy in the past, the clinician should find out details about the type of treatment the child received and how the child responded. Did his articulation or language difficulties improve? Did his stuttering first appear or worsen during treatment? If so, the clinician should pay particular attention to indications that the child may think of himself as a poor speaker and may believe that speaking is difficult. The interviews and questionnaires I suggested in the previous section on feelings and attitudes will help you explore this possibility, and the therapy approaches described in the chapters on beginning- and intermediate-level stuttering are designed to help a child regain confidence in his ability to speak easily and well.
Other Factors
Other factors can influence the outcome of treatment. We recommend evaluating all factors that may have precipitated or are maintaining a child’s stuttering so that they may be included in the child’s overall treatment plan.
Physical Development
My main concern in this area is that motor development may be lagging behind language development. A child with speech-motor delays may benefit from therapy that helps him coordinate respiration, phonation, and articulation, thereby reducing stuttering. He may also benefit from procedures that help him learn to stutter easily and openly, rather than becoming tense and frustrated if his fluency breaks down under stress. Such children’s treatment should also focus on building self-esteem, which may be low in children who are not well coordinated. One way to help build a child’s confidence is to figure out what he is good at or what he would like to improve his mastery of and encourage that. I sometimes make up games, like tossing a ball into a waste basket at greater and greater distances and have the child practice this during breaks from working on speech. Most children will delight in improving their skill at tasks at which they have some success, especially when they can do better than the clinician.
Cognitive Development
I try to find out whether or not cognitive stresses of school may be increasing the general demands experienced by the child, which I’ll discuss further in a following section on academic adjustment. If a child has academic difficulty or a learning disability, we may need to adjust our approach to treatment to make sure that he understands our explanations and examples.
Social-Emotional Development
I am interested in how well a child fits in with his classmates, how comfortable he feels about talking and relating to others, and how often he feels a need to hide his stuttering. Some children are friendly and outgoing even though they stutter and are supported by their classmates. These social skills are a positive factor in their prognoses for recovery from stuttering. Other children may be sensitive or selfconscious, and stuttering compounds their self-concern and keeps them from relating easily to others. Such children need help in relating more easily to their classmates. Evaluation of this component can be accomplished through teacher, parent, and child interviews; classroom observation may be helpful, too.
I am also concerned with the extent to which a child’s home environment provides support and security. This information comes primarily from parent and child interviews. Parents often provide insight into conditions surrounding the onset of stuttering and conditions under which it gets better or worse. I sift through this information, and with the parents’ help, determine whether something can be done to improve the child’s self-esteem. For some children, school psychologists have been helpful in building self-esteem and helping them improve their social adjustment.
Academic Adjustment
Parent, child, and teacher interviews allow me to find out how well the child is doing in school and how much he likes it. Stuttering may appear for the first time or worsen when a child is under the stress of learning many new things. For example, reading aloud in class when just learning to read is likely to put substantial demands on a child’s resources for language formulation and speech production. The child must make “second-order mappings of meanings and lexical units from speech” (Gibson, 1972) while simultaneously translating the written representation into units appropriate for speech production. Thus, some academic challenges may be more demanding for a child who stutters, and his stuttering in school should be understood in relation to this. In practical terms, clinicians can determine if a child needs extra help in certain academic areas through discussions with his teachers about which speaking situations in school are most difficult for him. If the child has more difficulty in certain academic situations, these should be given extra attention in treatment when planning generalization of more fluent speech.
At this point in a clinic-based evaluation, the clinician pulls together the information collected from the case history; parent, teacher, and child interviews; speech samples; and classroom observations. This information helps the clinician determine the developmental/treatment level of the child’s stuttering, which will give direction for treatment.
Most school children are at beginning or intermediate levels of stuttering. Beginning-level stuttering is characterized by physical tension, hurry, escape behaviors, awareness of difficulty, and feelings of frustration. The intermediate level also involves tension, hurry, escape behaviors, and frustration, as well as avoidance behaviors as a result of fear and anticipation of stuttering. In addition to a child’s stuttering behaviors and feelings, current developmental and environmental pressures must be considered in planning treatment. Such pressures can be uncovered from parent, teacher, and child interviews and the speech sample. Some pressures may result from other speech and language disorders, motor problems, or pressures in the child’s home. Goals can be formulated with the parents’ input for alleviating those pressures that can be changed and helping the child cope with those that can’t be changed. Some pressures can be dealt with in treatment, but others may require parent counseling or referral to other professionals.
Closing Interview
The closing interview provides an opportunity to summarize my immediate impressions for the parents and make recommendations about treatment. It also provides an opportunity to discuss the crucial role parents can play in reducing environmental pressures. I point out the many beneficial things they have done about their child’s speech and assure them that stuttering was not caused by anything they have done. Although some parents may have created conditions in which a child’s predisposition to stutter has been transformed into a serious problem, it does not help to make an issue of this. Rather, we want to convince them that they are in a key position to help.
After describing clearly and simply what I observed about the child’s stuttering, I summarize my thinking about appropriate treatment. I do this in only general terms because parents’ main concerns at this time are not the details of treatment but the prospects for their child’s future. Therefore, I rely on my experience to describe likely outcomes. For example, I might say that a combination of many factors will determine the child’s outcome. These include the natural increases in fluency that occur as a child matures, feelings of self-acceptance that a child develops when he finds that people accept him whether or not he has trouble with his speech, and his learning ways to speak more fluently. When I talk about the child’s prognosis, I always include some aspect of the parents’ role, such as their acceptance of the child’s speech or their participation in treatment, as part of the formula for recovery. Sometimes, a key aspect of the parents’ acceptance of stuttering is realizing they are not responsible for curing it. If I feel that there is a good chance the child will have some stuttering remaining after therapy, I talk with the parents about this possibility, indicating that many people who have some stuttering remaining lead highly successful lives. A few who come to mind are Malcolm Fraser, a highly successful businessman who founded the Stuttering Foundation; Carly Simon, the singer; and Vice President Joe Biden.
After summarizing my impressions and describing some of the ingredients for recovery, I discuss some of the things the parents can do to promote recovery. Specific suggestions depend on findings from our interviews, but the sections on parent counseling in the chapter on treatment of stuttering in school-age children present general ideas for parents’ involvement. Discussion of the family’s involvement in therapy is the most important part of the closing interview and in fact may continue for several more meetings. If I treat the child directly and in a clinic rather than in a school setting, I meet with parents weekly as part of treatment. In these meetings, I continue to help them explore how various changes in the home environment can facilitate their child’s fluency.
Public School Setting
The sequence dictated by IDEA after a referral for stuttering is made is for the multidisciplinary team to develop an assessment plan and carry it out. After the clinician gathers information using tests, observations, and interviews, she writes a report describing the affective, behavioral, and cognitive aspects of the child’s stuttering and his current performance in academic, nonacademic, and extracurricular activities. The assessment report should be brief (e.g., two pages) and understandable by lay readers, such as the child’s parents.
Then, an Individualized Education Program (IEP) team is appointed to consider the report and other information. The IEP team must decide the two issues mentioned earlier–does the child meet the state’s eligibility standards, and does the child’s stuttering have an adverse effect on his education? The first of these issues is resolved by the data the clinician gathered, particularly the information about the frequency and severity of the child’s stuttering and his feelings and attitudes. The second issue is usually more complex because it is necessary to show that the child does not perform as well as he might in school because of his stuttering. For example, the team can conclude that his stuttering prevents him from participating as much as he might in class or in extracurricular activities with his peers. Evidence of this can be obtained from measures of communicative functioning in school, such as the Teacher’s Assessment of Student Communication Competency (Fig. 8.9) (Smith, McCauley, & Guitar, 2000), observations of the child in the school, and interviews with teachers and parents. Lisa Scott Trautman (personal communication, July 30, 2003) noted that adverse effects can be shown by demonstrating that the child cannot meet the school district’s curriculum objectives because of his stuttering. Examples of such objectives might be that students will be active in class discussions or that students must be able to speak effectively in front of a group.
If the evaluation determines that the student is eligible for services, an IEP team, often headed by the SLP, develops measurable goals and short-term objectives (also called “benchmarks”) as well as services to be provided that will help the student improve his performance in all aspects of the educational setting. These goals and objectives are considered in detail in the chapters on treatment of beginning and intermediate stuttering.
The assessment of an adolescent or adult can be very rewarding for a stuttering clinician. Although some clients may be discouraged because they have been stuttering for years and skeptical that you can really help them, many are highly motivated to work on their speech and ready to begin this work during the evaluation. Your challenge is to take advantage of this motivation, get them working immediately, and to give them realistic hope that hard work and resolve can change their speech and maybe their lives.
Clinic Versus School Assessment
This section is written as though the evaluation is being carried out in a clinic rather than a school. When the setting is a public school, the evaluation process is determined by the Individuals with Disabilities Education Act (IDEA ’97) and the laws of each state. The guidelines for this process were described in the previous section on evaluating a school-age child. For the adolescent, an additional consideration is his participation in the IEP process and his transition beyond high school. When a student reaches age 14, his input is sought by the IEP team, and he gradually becomes an active member of the team, not only with regard to his present situation, but also in terms of his aspirations beyond secondary school. When a student reaches age 16, transition plans are a mandated part of the IEP. At age 18, students take over responsibility from their parents for signing off on documentation.
Case History Form
A case history form is sent to adult clients (those over age 18 years and beyond high school) several weeks before their appointment. A copy of this form is shown in Figure 9.4. Because adolescents are often seen in schools, the clinician encourages them to fill out the form with help from their parents for parts of it.
This form requests information that would be appropriate for most speech-language disorders and can be used with all adult clients referred for speech or language problems. It also allows the clinician to learn ahead of time whether the client referred for stuttering may have a different or additional disorder. The form also gives the clinician information about the extent to which stuttering, if that is the problem, affects a client’s life.

Figure 9.4 Case history from – adult and adolescents
Attitude Questionnaires
I assess clients’ communication attitudes through observations, interview questions, and questionnaires. Because I want to be able to analyze completed questionnaires before the diagnostic interview, I prefer to send them to clients and ask them to complete and return the questionnaires before the interview. If this is not possible, clients can complete them when they arrive for an evaluation before the initial interview or as a less desirable alternative, after the initial interview. Prior to the interview, follow-up questions based on information from the case history and questionnaires, which are described in the section on feelings and attitudes, can be prepared to further explore a client’s attitudes.
Audio/Video Recording
It is important to sample a client’s speech in several situations to get an adequate picture of his stuttering. I ask clients to video or audio record themselves talking in one or two different situations outside the clinic and get the recording to me prior to the evaluation. It is usually easy for clients to record themselves talking to someone on the phone, recording only their own voices and not the person on the other end of the line. Some clients can also record themselves talking face-to-face with a friend or family member. If I listen to the recording(s) before an evaluation, I am better prepared to understand the client’s stuttering and to plan various trial-therapy strategies.
I begin by welcoming the client and reviewing the procedures I will use to evaluate his problem, such as interviewing him about his stuttering and his feelings and attitudes, video recording his speaking and reading, examining what he does when he stutters, and trying to determine if he can change it. I let him know that after the initial part of his evaluation, I will ask him to wait while I analyze the information I’ve obtained before meeting with him to share my findings and recommendations. If there are any forms or questionnaires I haven’t already obtained from him, I’ll have him complete those while I analyze the other data.
I begin our interview with an open-ended question such as, “Tell me the problem that brings you here today,” or “Why don’t you tell me about your stuttering?” The first question might be used if I don’t know what is motivating the client to come for an evaluation at this time; the second question I use when I already know from prior information why the client has come right now.
Once a client has had a chance to describe his speech problem, I ask further questions to try to get a deeper understanding. The following are typical questions that I ask with a brief commentary about why I’m asking them. Sometimes, I group several questions together (e.g., a question to start the client talking about a particular topic and follow-up questions that I ask if the first question doesn’t elicit all of the desired information). I ask only one question at a time, listen carefully to the client’s response and try to understand the client’s underlying feelings.
1. When did you begin to stutter? How has the way you stutter changed over the years?
I realize that in answering the first part of this question, a client may just be reporting what parents told him about his stuttering. The accuracy of his response may be questionable, but at least I’ll learn his perception of the onset. The second part of the question–about changes over the years–may reveal what kinds of things affect the way a client stutters. Does he stutter more severely because of a recent job change or a threat to his self-esteem, such as a divorce or loss of employment? Less frequently, I may find out that a client began to stutter in late adolescence or as an adult. If so, I would want to consider the possibility of neurogenic or psychogenic stuttering, which is discussed briefly in the upcoming section on diagnosis and more fully in Chapter 15, on other fluency disorders.
2. What do you believe caused you to stutter?
This may give some insights about a client’s motivation. For example, a woman whose speech I once evaluated reported that her mother and several brothers stuttered and that her stuttering was therefore a genetic problem that could not be helped. This led us to confront the issue of whether or not she was likely to change.
In addition, I sometimes find that clients have misinformation about possible causes of their stuttering. If I can give them more appropriate information, their attitudes about the problem may change, and their motivation may increase. I have met individuals who come to the evaluation believing that their problem is entirely psychological. After I discuss current views of stuttering, they are relieved to know that they are likely able to modify their speech without longterm psychotherapy.
3. Does anyone else in your family stutter?
I might find that a parent stutters, which can be significant because a parent’s attitudes about his or her own stuttering may have had a profound effect on the client. Moreover, knowing about other family members who stutter and how they have responded to it may provide a better understanding of the factors related to this client’s stuttering, which may be useful in treatment. For example, someone I am interviewing may have had a parent who stuttered but who never talked about it. I might then want to explore whether the individual I am working with feels especially ashamed of his stuttering, feels it gives him an important bond with the parent, or both.
4. Have you ever had therapy for your stuttering? What did the therapy consist of? How effective do you think it was?
This information is important in planning therapy. For example, if a client had received a type of therapy that he felt did not help, it would be unwise to use that type of therapy with this client. But if a client has had success with therapy but has regressed slightly or moved away before treatment was finished, using this type of therapy again may be most appropriate. It is important that clinicians be familiar with various types of therapy that clients may have undergone. Most current therapies emphasize either modifying stuttering behaviors (stuttering modification) or learning to talk in ways that eliminate stuttering (teaching fluency skills).
5. Has your stuttering changed or caused you more problems recently? Why did you come in for help at the present time?
Responses to these questions allow clinicians to see the current problems faced by the client and also obtain some inkling of the client’s motivation. For example, a client may have been offered a promotion if he can improve his speech or may have recently learned of the clinic’s treatment program and is hoping for some relief from a long-standing problem. The following four questions about the client’s pattern of stuttering are closely related to one another.
6. Are there times or situations when you stutter more? Less? What are they?
7. Do you avoid certain speaking situations in which you expect to stutter? If so, which ones?
8. Do you avoid certain words on which you expect to stutter? Do you substitute one word for another if you expect to stutter? Do you talk around words or topics so you won’t stutter?
9. Do you use any “tricks” to get words out? Escape behaviors?
These four questions will provide information that is useful in planning therapy because they tell us something about the client’s most difficult situations, how he feels about them, and how he deals with them. This information may also corroborate what has been learned from the questionnaires that the client completed and will also reveal how aware he is of his stuttering behaviors.
10. Have your academic or vocational choices or performance been affected because you stutter? How?
The client’s answers can be used to help plan later stages of treatment in which new behaviors and new challenges are attempted. They may also prompt the clinician to refer clients in later stages of treatment to an academic or vocational counselor to help them make more appropriate choices for themselves.
11. Have your relationships with people been affected because you stutter?
As with question 10, I can use this information to plan a client’s hierarchy of generalization, moving from easy to difficult social situations gradually if the client finds social interactions difficult. I also need to know how much a client blames his stuttering for any of the difficulties he has in social interactions. A client may be socially inhibited because he is sensitive and vulnerable to expected listener reactions. Such sensitivity can be assessed by observing his facial expressions and body movements while stuttering as indicators of affect. If he appears to be relatively unaffected emotionally by his stuttering but professes to have difficulty relating to people, he may benefit from counseling or psychotherapy that focuses on resolving this interpersonal difficulty.
The decision to refer an individual for psychotherapy as an adjunct to stuttering therapy can seldom be made in the evaluation session. It may be that a few therapy sessions are needed to learn more about a person and to develop the client’s trust before a successful referral can be made. If psychotherapy is recommended too hastily, a client may believe that I think his stuttering is too great a problem for me to handle, perhaps an insurmountable problem or one that I secretly believe is due to a psychological disorder. However, if I work with him and he starts making some progress before I refer, he will likely feel supported and may be more likely to benefit from psychotherapy.
12. What are your feelings or attitudes toward your stuttering? What do youthink other people think about your stuttering?
A client’s responses will be used to help determine some of the foci of treatment, such as desensitization procedures to decrease fear as well as shame or guilt about stuttering. Perceptions about others’ views of his stuttering may need to be confronted with various “reality-testing” tasks to find out what people really think.
13. What are your family’s (parents’, spouse’s, children’s) feelings, attitudes,and reactions toward your stuttering and toward the prospect of your being in therapy?
This information can identify sources that may positively or negatively affect a client’s motivation and may be an important consideration in planning therapy.
14. Is there anything else that you think we ought to know about your stuttering?
This gives the client a chance to get anything off his chest that he may be holding back or an opportunity to discuss issues that occurred to him only after other questions were asked.
15. Do you have any questions you’d like to ask me?
Sometimes an adolescent or adult has questions about stuttering that he has been reluctant to ask, and this may give the clinician an opportunity to answer them. On the other hand, a client may want to ask about the length and type of treatment or other issues that are best dealt with after his assessment is completed. In this case, the clinician explains why she needs to delay responding but will keep the questions in mind to answer during the closing interview.
Speech Sample
In this part of the evaluation, the client’s overt stuttering behaviors are assessed. Although I always video record the entire evaluation, if the client has given permission, I pay particular attention to the recording of this section because I will need to analyze it carefully afterward. Clinicians use a variety of procedures for assessing overt stuttering. Next, I shall describe in detail the tool I currently use and then note other available options.
Stuttering Severity
As indicated in the previous chapter, stuttering severity is usually measured using the Stuttering Severity Instrument (SSI-4), a moderately reliable tool that is commonly used to assess the severity of stuttering. To obtain appropriate samples, I have the client talk about a familiar topic, such as his work, school, hobbies, vacations, sports, or entertainment. It is important to get about at least 300 syllables of the client’s talking, so five or 10 minutes are usually enough, depending on the client’s fluency. Then, I provide the client material at an appropriate reading level, such as the passages in the SSI-4 examiner’s manual, and ask him to read aloud for about three minutes to get 200 or more syllables of reading.
As I noted earlier, we often gather more than one sample of spontaneous speech from adults and adolescents. A sample of speech during a telephone conversation in the clinic can be video recorded and scored using the SSI-4. In addition, I use samples the client has brought or sent in. If the sample from another environment is audio recorded rather than video recorded, I score it for both frequency of stuttering and speech rate, as described below.
Other Measures of Stuttering
If I am assessing stuttering many times throughout the course of treatment or assessing samples that I cannot visually analyze, such as those audio recorded by a client in his natural environment, I use a combination of frequency of stuttering (percentage of syllables stuttered) and speech rate (syllables spoken per minute). These measures, which were first described in Andrews and Ingham (1971), together require much less time than the SSI.
Starkweather (1991) has presented a case for capturing the amount of time that stuttering takes. This is done by totaling the duration of all disfluencies and pauses in a sample and dividing this total by the overall time spent in speaking, thereby giving the clinician a measure of how much an individual’s stuttering interferes with the rate with which he can communicate information.
As part of a determined effort to improve the reliability of stuttering measures,
Ingham and his colleagues (Ingham, Cordes, & Gow, 1993; Ingham, Cordes, & Finn, 1993) developed a time interval system of assessment. They have shown that when judges determine whether or not four-second intervals of continuous speech contain one or more stutters, interjudge reliability is higher than when moments of stuttering are counted; however, the clinical usefulness of this procedure has not been determined.
Speech Rate
In addition to measuring stuttering severity using the SSI, I also assess a client’s speech rate. I believe, as many other clinicians do, that speaking rate often reflects the severity of stuttering, as well as its effect on communication. If a client’s speech rate is markedly slower than normal, communication may be difficult for him. A description of the procedure for measuring speech rate was given in Chapter 8, “Preliminaries to Assessment.”
Normal speaking rates of adults range from around 115 to 165 words per minute, or about 162 to 230 syllables per minute, with a mean of 196 syllables per minute (Andrews & Ingham, 1971). Adults’ normal rates for reading aloud are faster, ranging from about 150 to 190 words per minute (Darley & Spriestersbach, 1978), or about 210 to 265 syllables per minute (Andrews & Ingham, 1971).
Pattern of Disfluencies
Throughout my evaluation of adult or adolescent stutterers, I observe the client’s patterns of stuttering. For example, I try to roughly determine the proportions of core behaviors that are repetitions, prolongations, or blocks and ask myself a number of questions about the client’s stuttering. During blocks, where and how does he shut off airflow or voicing? What are his escape and avoidance behaviors? Does he end the stutters quickly with pushing and tension? Is he able to tolerate being in blocks, or does he speak in unusual or vague ways to avoid stuttering? More details on various escape and avoidance patterns can be found in Chapter 7 on the Development of Stuttering.
As I explore the behaviors that constitute a client’s stuttering, I comment on them, question him about how typical this sample of his stuttering is, and ask about the escape and avoidance behaviors we’ve observed. If a client doesn’t seem too uncomfortable confronting his stuttering, I ask him to teach me how to stutter like he does, and we work together, with both the client and myself emulating his various types of stuttering. This does not need to be an exhaustive exploration, because I will do much more in treatment. Here, I am trying to accomplish three tasks: (1) model an “approach” rather than “avoidance” attitude toward stuttering, showing calmness and objectivity about behaviors that the client may feel are shameful and perhaps even terrifying; (2) study the client’s emotional reaction when he comes face-to-face with his stuttering and perhaps reduce some of his fear; and (3) teach both of us about what the client does when he stutters so that he can begin to learn how to change it.
Trial Therapy
I try therapy techniques with clients during their assessment sessions for several reasons. First, I try to get an idea of how a client responds to different therapy approaches, which provides me with information I may use in talking with him about possible treatments. Second, trial therapy can help me to make a differential diagnosis between developmental stuttering and stuttering with a neurological or psychological etiology. Third, it gives clients a preview of things to come and provides them with hope and motivation to follow through on treatment.
I begin by asking a client to modify his stuttering, which can be done easily in the context of studying his patterns of disfluency, as described in the preceding section. In fact, this exploration of stuttering with a client is a condensed version of the first stage of treatment that aims to change stuttering to an easier pattern. Once a client is able to emulate his stuttering to a small degree, I carry out trial therapy by coaching him through the following sequence:
1. First I encourage him to stutter, telling him we must have a sample of the behaviorwe are trying to change. Then I ask him to “freeze” during a moment of stuttering but maintain the level of physical tension and posture of his stuttering as I encourage him to stay in the moment of stuttering. In other words, I ask him to catch a stutter and prolong it. This may require modeling of how to hold a moment of stuttering right on the sound that’s being stuttered. This is easier with a continuant sound, like /m/, but will be harder and need more coaching for plosives, such as /b/. You will probably have to model for the client how to prolong the posture required for holding a stutter on a plosive. It is key that you and he identify the exact posture that is associated with the moment of stuttering and hold onto it. It is important that during this activity the clinician praises the client enthusiastically for catching and holding onto a moment of stuttering. This helps the counterconditioning process–pairing a positive stimulus with a behavior about which the client feels negative.
2. Have him become aware of what he is doing that creates the stutter. For example,where is he holding back sound or airflow? Lips? Tongue? Larynx? All three? As you are helping him explore what he’s doing when he stutters, use plenty of praise for being able to stay in the stutter. This is an experience charged with fear and frustration for most adolescent and adult stutterers. They need the counterconditioning that you provide by rewarding their maintaining of this stuttering moment. It’s similar to treatment for a phobia: staying in contact with a feared object (a spider or even, in some cases, a rabbit) reduces fear if there is reward provided by another person. It has been said that the physical awareness of what the client is doing with his body as he stutters can be an antidote to the fear he might otherwise be feeling (Zebrowski, personal communication, October 18, 2011).
3. Have the client change his behaviors that are maintaining stutters by (a) releasingexcess physical tension wherever he can feel it, (b) starting to move rigidly held structures, (c) getting voicing or airflow going, and/or (d) allowing himself to breathe. I may stop a client’s trial therapy here if he is unable to release physical tension or does it only with obvious difficulty.
If a client seems able to make these changes easily, I go one step further. I ask him to hold on to the stutter, which has become voluntary by now, and to prolong the airflow or voicing for several seconds (while I tell him how great it is that he can do this!), and then produce the remainder of the word slowly. (Some of my clients call this “catch and release.”) If a client is able to do this with coaching, I ask him to do it while reading without my coaching. This is enough. No matter how much or how little our client is able to do, I want to stop when he is feeling successful.
Another approach to trial therapy is to change the client’s habitual way of talking so that stuttering is decreased substantially or prevented. I begin by reducing my own speech rate as I describe the aim of this exercise to the client, which is to produce words very, very slowly. I use a written sentence that begins with a vowel or a glide, going over it word by word, teaching the client to use gradual and gentle onsets of voicing and to stretch each sound, whether vowel or consonant. This is essentially the “prolonged speech” or “fluency-facilitating targets” used by some fluency-shaping approaches, such as the Camperdown Program (O’Brian, Packman, & Onslow, 2010) and the Fluency Plus Program (Kroll & Scott-Sulsky, 2010). The clinician needs to provide a good model for each word and to give feedback frequently. When words are produced slowly enough with each part of the speech production system (respiration, phonation, and articulation) moving in slow motion and without excess tension, then fluency results. After a client is able to produce each word of the sentence in this way, he is then coached to produce the entire sentence, linking each word to the next. Breath supply should be monitored closely, so that pauses for breath are taken whenever the client would take a breath naturally. Again, accurate modeling and frequent feedback are crucial at earlier stages of treatment.
As an example, the sentence, “Apples are a red fruit,” should take from 15 to 20 seconds to produce, with a pause for a new breath after the word “a.” The /p/ in “Apples,” the /d/ in “red,” and the /t/ in “fruit,” each should be produced without stopping airflow, making these plosives sound like fricatives. If clients are particularly adept at this, they can be taken all the way to saying short sentences in conversational speech that are produced in this slow, fluent manner. However, clients who have difficulty should be coached only through the production of the short, written sentence, and care should be taken to stop this activity before they experience failure.
Feelings and Attitudes
A variety of questionnaires can be used to assess various aspects of a stutterer’s feelings and attitudes about communication and stuttering. In Chapter 8, I described those questionnaires that I use regularly. These include the Modified Erickson Scale of Communication Attitudes (S-24) (Andrews & Cutler, 1974), the Stutterer’s SelfRating of Reactions to Speech Situations (Johnson, Darley, & Spriestersbach, 1952), the Perceptions of Stuttering Inventory (Woolf, 1967), and the Locus of Control of Behavior Scale (Craig, Franklin, & Andrews, 1984).
Other Speech and Language Behaviors
As I interact with a client during the interview, I informally assess his comprehension and production of language, his articulation, and his voice. I also screen his hearing. If I suspect that there may be an articulation, language, voice, or hearing problem, I follow up with further evaluations. Adolescent language assessment procedures can be found in Nippold (2007) and Nelson (1998), and procedures for assessment of articulation can be found in Bernthal, Bankson, & Flipsen (2009). I let a client’s concern about other disorders guide us in treatment. If, as I have found occasionally, a stuttering client also produces distorted /s/ or /r/ sounds, I discuss it with him. If he is not concerned, I don’t believe it is necessary to treat that problem. However, if I believe that an articulation, language, or other problem handicaps a client communicatively, I advise treatment for that problem also. Sometimes I deal with voice problems differently. I have found that some stutterers may be hoarse, but I suspect this may be the result of laryngeal tension related to stuttering. If stuttering treatment is successful, hoarseness may disappear. Again, I take my cue from the client. If the problem bothers him and isn’t remediated by treatment, I address it. If hoarseness is of recent origin and not associated with a cold, I may refer him for an otolaryngological examination to rule out serious laryngeal pathology.
Other Factors
In this section, I discuss the evaluation of the following factors: intelligence, academic adjustment, psychological adjustment, and vocational adjustment. Each of these factors can affect the treatment of an adult or adolescent stutterer and therefore must be considered in planning therapy. The factors are considered briefly here, but some are covered in depth in the chapter on other fluency disorders (Chapter 15).
If a client has below-normal intelligence, he may have difficulty following the regimen of a typical therapy program. Usually, clinicians know beforehand if a client scheduled for an evaluation has Down Syndrome or some other condition characterized by below-normal intelligence. An adolescent stutterer who has been identified as developmentally delayed will likely already be receiving special education. Adults, too, are usually identified as mentally handicapped if this is the case, and either the referral source will report this, or a guardian will have filled out the case history form.
Problems of academic adjustment in an adolescent who stutters usually become apparent from either the original referral or interviews with the child’s teachers as part of the evaluation process. These interviews are described in more detail in the section on the school-age child. An example of poor academic adjustment relevant to stuttering could be a student’s conflict with a teacher who insists on oral presentations that the student is unwilling to do. The IDEA 1997 process mandates a team approach to solving such problems. This process will be described in the chapters on treatment.
The research reviewed in Chapters 2 and 3 suggests that there are no group differences in the psychological health of stutterers and nonstutterers. However, we sometimes see individuals who stutter who do not function well in their environment. They may be unable to achieve a satisfying marriage, unable to hold a job, or are socially withdrawn. Clinicians need to be alert to the effects that adjustment problems may have on treatment. If psychological problems are suspected of interfering with treatment progress, the clinician may wish to refer the client for a psychological evaluation. In such cases, the clinician should take care to ask professional colleagues for recommendations of the most effective psychotherapists in the area.
Psychological problems that are relevant to stuttering also may become apparent during the interview when the onset of stuttering is explored. Sudden onset after a psychological trauma, particularly if onset is in late adolescence or adulthood, may indicate psychogenic stuttering. I have found that if the psychological effects of the trauma have subsided, an adolescent or adult client may respond well to the integrated approach to treatment described in Chapter 14. If it is clear that psychological factors are still affecting the client’s speech and behavior or if there is doubt, I refer the client for a psychological evaluation. Unless the disorder is a psychosis, in which case stuttering therapy may not be recommended, clients with psychological problems may respond well to a combination of psychotherapy and stuttering therapy.
Interview with Parents of Adolescent
When I evaluate an adolescent who stutters, I also talk with his parents, sometimes separately from the adolescent, to obtain more background information about the student, to give them an opportunity to express their concerns and feelings privately, and to tell them about the evaluation process and the options for treatment.
I begin the interview by asking the parents to describe the problem as they see it and encourage them to express their fears, concerns, and frustrations, as I listen carefully. I try to get an understanding of how their child functions within the family and usually ask such questions as: “What is his stuttering like at home?” “How does he seem to feel about it–is he embarrassed or does he show fear of talking or anger about his speech?” “How do you feel about it?” “What are your and other family members’ reactions to it?” “What do you do when he stutters?” “Has he been seen anywhere else for therapy?” “If so, what were the results?” Even though I am putting some of these questions in groups, I am careful to ask one question at a time and listen carefully to the answer before I ask another question. Although parents may ask what can be done to help their child and what they should do, I prefer to wait until after I have seen the youngster before answering these questions.
Adolescents strive to become more and more independent of their parents, and I have found that therapy works best if an adolescent is treated as an adult. I begin fostering independence by talking first to teenage clients separately from their parents so that they can give me their own views of the situation and how they view the prospect of treatment. After this and after my meeting with the parents, I meet with the parents and teenage clients together to seek mutual agreement about their respective roles in treatment. This is often an important time. It serves to let teens know that I respect their ability to work independently from the parents, and it serves to let the parents know that they can be most helpful by being supportive but not directive.
After I gather the information just described, I need to determine whether the client stutters and if so, what treatment level is appropriate. Typically, teenage clients are advanced stutterers; however, some are still in the intermediate stage. But first, let us consider the possibility that a teenager turns out not to be a stutterer.
In rare cases, teens who are normally but highly disfluent may be referred by teachers, employers, or friends. Most have phrase repetitions, circumlocutions, revisions, and hesitations, which are the types of disfluencies described in Chapter 7 as normal. Such disfluencies are observed relatively infrequently after children’s elementary school years; however, some adolescents and adults may simply be at the disfluent end of the continuum of normal fluency. In addition to the differences in type and number of disfluencies, secondary behaviors and negative feelings and attitudes will be absent. Our role in such cases is to explain to the individual and to the referring person (if this is a referral) that this kind of speech is not abnormal and need not be of concern. It may also be emphasized to the referring source that excessive attention to these disfluencies may be more harmful than helpful. If the client or referring person feels strongly that the disfluent speech interferes with communication, a fluencyoriented treatment described in the chapters on treating intermediate and advanced stuttering may be offered to the client.
Another need for differential diagnosis, in addition to identifying cases of normal disfluency, is ensuring that cluttering, neurogenic disfluency, and psychogenic disfluency be identified and distinguished from “typical” or “developmental” stuttering. Moreover, it is also necessary to rule out disfluencies caused by wordfinding difficulties that we might find in a person with a learning disability.
Some of the salient features of cluttering in adults and adolescents are rapid, sometimes unintelligible speech, frequent repetitions of syllables, words, or phrases, lack of awareness or concern about their speech, disorganized thought processes, and language problems. Cluttering often coexists with stuttering, and both disorders may respond to a highly structured, fluency-shaping approach for treatment. Evaluation and treatment procedures for cluttering are described in Chapter 15.
Neurogenic disfluency in adolescents or adults is usually the result of stroke, head trauma, or neurological disease. Symptoms are likely to be repetitive disfluencies but may include blockages as well. Because stuttering commonly begins in childhood, if a client reports onset of stuttering after age 12, a neurogenic-based disorder is a possibility. In almost all such cases, onsets of neurogenic-based fluency problems are clearly linked to a well-defined episode of neurological damage. A section of Chapter 15 is devoted to evaluation and treatment of neurogenic stuttering.
Disfluency that begins in adolescence or adulthood can also result from psychological trauma. When late-onset disfluencies are seen that are associated with psychological stress and conflict or the onset of a psychiatric condition, psychogenic disfluency should be suspected. Traditional treatments, such as those described in the chapters on treatment of intermediate and advanced stuttering, may or may not be helpful. The patient should be referred for both psychological and neurological assessments, so that treatment needed in these areas will be identified and provided. See Chapter 15 for more information.
When a clinician determines that stuttering treatment would be appropriate for a client, whether the stuttering had a typical onset during early childhood or has another etiology, the focus turns to a consideration of what level of treatment to select for the client. As I indicated earlier, adult and adolescent stutterers are most likely to be at advanced developmental and treatment levels. Signs of this level include the core behaviors of repetitions, prolongations, and blocks, all with tension; the secondary behaviors of escape and avoidance; and negative feelings and attitudes about communication in general and stuttering in particular.
Determining Developmental and Treatment Level
The determination of a developmental/treatment level for an adolescent or adult stutterer is based largely on the client’s age. Intermediate and advanced treatment approaches are well suited for clients whose core behaviors are blocks, who have escape and avoidance behaviors as secondary symptoms, and whose attitudes about speech are relatively negative. A client suited to the advanced-level treatment will usually have more entrenched negative attitudes about speech and himself as a speaker simply because he has been stuttering longer. It’s possible that someone who is at the advanced level will have developed an extensive repertoire of avoidance behaviors so that actual stuttering behaviors are rare, but the individual’s life is highly constrained by his efforts to avoid and hide his stuttering. The major difference between intermediate and advanced treatment levels is that more independence and responsibility are required of clients at the advanced level. Consequently, clinicians ordinarily place adult clients at the advanced level but determine an adolescent’s placement based on how much responsibility he can take for self-therapy.
Intermediate Stuttering
A client whose stuttering is at the intermediate level will probably be younger than mid-teens. His stuttering pattern will be characterized by escape and avoidance behaviors and considerable tension on blocks, prolongations, and repetitions. He will also be avoiding some speaking situations. Moreover, his feelings and attitudes, as revealed in questionnaires and interviews with him and with his parents and teachers, will suggest many negative speech attitudes.
Advanced Stuttering
Individuals who fit into the advanced developmental/treatment level are well into their teens and sufficiently mature to handle the assignments used in advanced treatment. Their stuttering pattern is similar to the intermediate stutterer’s, but their patterns of avoidance and escape may be more habituated (i.e., patterns appear to be highly automatized and rapidly performed). They will probably avoid difficult speaking situations whenever possible, and I often find strong negative self-concepts and negative anticipations of listener reactions as well. An advanced stutterer may feel, for example, “I must be awfully incompetent to talk like this,” or “People think I’m dumb because I stutter.”
Closing Interview
I will assume here that the client is a person with developmental stuttering rather than another type of fluency disorder. By this point in the evaluation, I have a pretty good picture of his stuttering and how I will start therapy. I begin by summarizing my impression of his stuttering pattern (i.e., core and secondary behaviors) and his attitudes and feelings. One of my aims is to let him know I have some understanding of his stuttering and why he does what he does when he stutters. I feel it is important to let him know that, given his level of stuttering, it is no surprise that he would use the various secondary behaviors and avoidance tactics that he does. I accept these behaviors rather than criticize them and let him know that I feel I can work with him and help him discover other ways to respond. I try to ensure that he feels he will not be alone, that I will be working alongside him, and that I will gradually give him more and more responsibility to work on his own.
Then, I briefly describe some therapy options and discuss the possibilities with him. With my guidance, the client and I decide on a treatment approach. Afterward, I give him some written suggestions to begin the process of his taking responsibility for part of his treatment. This will also take advantage of the fact that, as indicated earlier, many clients are highly motivated to change at the time they come for an evaluation. I may not do this with adolescents, but there are exceptions. Some adolescent clients are reluctant to participate in therapy rather than being highly motivated because of their desire to close ranks with their peers and distance themselves from adults. With adolescents, I often end our evaluation session by striking a bargain to try at least four sessions of therapy before they make a decision about treatment. I may also give them the booklet, Do You Stutter: A Guide for Teens (Fraser & Perkins, 1987), and the video, Do You Stutter: Straight Talk for Teens (Guitar & Guitar, 2003), both available from the Stuttering Foundation. These items will help them learn about therapy and develop realistic and motivating expectations about its potential outcome.
If the client has few avoidances and relatively mild stuttering, I am likely to start treatment with fluency shaping. If the client’s stuttering is moderate to severe and/or he has relatively many avoidances and fears, I am likely to start with stuttering modification treatment. See the next chapter for an overview of both approaches. An exception that I may make is when a client has many fears and avoidances, but seems unwilling to confront them. I may begin working with this client using fluency shaping. Some clients who are at first unwilling to “touch the hot stove” of stuttering, will be able to confront and change their stuttering if they first get some fluency through fluency shaping.
At the end of the closing interview, I ask a client if he has any questions about the evaluation. I also try to answer the questions asked in the initial interview that I postponed for response until after the evaluation. Adults and adolescents sometimes ask how long treatment will take. This is a reasonable question, given that they need to budget time and money to undertake treatment, but I have no easy answer for this difficult question. With appropriate cautions about individual differences and unexpected issues, I reply that with hard work and a willingness to tackle difficult situations and to confront fears with my help, I believe that considerable progress can be made within a year of the onset of treatment.

• In evaluating a client who may stutter, your task is to decide
(1) if his disfluencies warrant treatment,
(2) if they do, you should also find out more about his history, current environment,speech behaviors, and reactions, as well as
(3) what treatment seems reasonable given these findings?
• In assessing a preschool child, the important questions to answer are
(1) whether the child is stuttering or is normally disfluent,
(2) what the probabilities are that he will recover without treatment, and
(3) if treatment is warranted, you need to determine if indirect (for borderline stuttering) or direct (for beginning stuttering) is best.
• It is important to obtain some information prior to the formal assessment. This includes a recording of the child’s speech at home and a completed case history.
• Key elements of the assessment for a preschool child are
(1) observation of parent-child interaction
(2) parent interview
(3) clinician-child interaction
(4) analysis of child’s speech
(5) screening of language, articulation, and voice
(6) determining risk factors
(7) deciding on the child’s need for treatment
(8) making follow-up recommendations to family
• In assessing a school-age child, the important questions are
(1) how supportive the parents are of the child’s problem
(2) how the stuttering is affecting the child’s performance in school
(3) how the child feels about his stuttering
(4) how motivated he is to work on it
(5) how supportive the child’s teachers are
• The assessment of the school-age child may proceed differently if he is being seen ina clinic or at school. If seen at school, the IDEA affects the process and mandates how assessment is carried out. If seen in a clinic, the clinician will have more contact with the family but needs to reach out to the school setting.
• Key elements of the assessment for a school-age child are
(1) initial contact and formal interview with the child’s parents
(2) interview with the child’s teachers
(3) interview with the child
(4) analysis of speech
(5) trial therapy
(6) assessment of other factors, including academic adjustment
(7) determination of appropriate treatment
• In assessing an adolescent or adult, the important questions are the client’s level ofmotivation and ability to carry out assignments independently, the severity of stuttering and degree of avoidance, the client’s feelings and attitudes about his stuttering, whether the problem is typical “developmental” stuttering or is cluttering, psychogenic, or neurogenic stuttering, and the appropriate type of treatment.
• Key elements of the assessment of an adolescent or adult are
(1) obtaining preliminary case history, attitude questionnaires, and recordings madeoutside of the clinic
(2) interviewing the client
(3) analyzing the client’s speech
(4) conducting trial therapy
(5) interviewing parents if client is an adolescent
(6) determining appropriate treatment
(7) summarizing findings and making recommendations in closing interview withclient.
• Whether the person is to be treated as a normally disfluent speaker or as someonewho stutters depends on your interpretation rather than a score. You must weigh what you see and hear to determine whether they indicate stuttering, normal disfluency, or even another disorder. From the flood of information you have gathered, you must extract the essential characteristics that support your choice of treatment.
• To hone your judgment, make evaluations a continuing process. The procedures Ihave suggested for assessment and diagnosis in this chapter will give you a good start, but stuttering is highly variable, and no individual can be completely evaluated in just an hour or two. Consequently, you will overlook an important element at times, and sometimes a vital clue will not be present in the samples of behavior you see during an initial evaluation. With good ongoing evaluation of a client, you will be able to change decisions and redirect therapy as additional information and understanding become available. You will also be able to evaluate the effectiveness of your treatment and improve it when needed.

1. How do you determine whether a preschool child is stuttering or is normallydisfluent?
2. Why is it useful to obtain audio/video recordings of a preschool child’s stutteringbefore the evaluation?
3. What are some indications that a parent of a preschool child who stutters feels sheor he is to blame? How can you help the parent deal with those feelings?
4. What do you tell the parent of a preschool or school-age child who asks you whatcauses stuttering?
5. What are the variables assessed in the speech of a preschooler to determine hisdevelopmental/treatment level?
6. What are the advantages and disadvantages of talking to a child about his stuttering?
7. Compare the involvement of the parent and the teacher in the evaluation of aschool-age child.
8. In what various ways do we assess the impact of the school environment on theschool-age child who stutters?
9. What are the benefits of obtaining both a reading and a conversation sample withschool children and adults?
10. In the section on evaluation of the adult and adolescent, what different pieces ofinformation that you may gather from the interview questions help you to assess the client’s motivation?
11. What are two reasons we suggest continuing evaluation after the initial assessment of clients who stutter?
12. Compare the assessment of the feelings and attitudes of a school-age child with the assessment in an adult.
13. What is the role of the Individualized Education Program team in the management of a school-age child who stutters?
14. What are the goals of trial therapy?
15. What are the major questions to be answered in the evaluation of an adult?
1. Role-play the part of a clinician in a parent interview, having a friend orclassmate play the part of a parent. Practice your listening skills by only listening and asking no questions as the “parent” describes in detail his or her child’s stuttering problem. Switch roles, then compare your impressions of the experience both as the parent and as the clinician.
2. Pair up with a friend or classmate who could pretend to stutter or with aperson who stutters and practice trial therapy that is appropriate for a schoolage child and then appropriate for an adult. Try both approaches, modifying stutters to make them less severe and modifying speech to produce fluency.
3. Pair up with a friend or classmate who doesn’t stutter, and have them talkrapidly about a complex topic so he or she produces normal disfluencies. See if they are able to “catch” their normal disfluencies and hold onto them (e.g., turn single repetitions into multiple repetitions or make prolongations longer). Can this be done with normal disfluencies? With only certain types of normal disfluencies?
4. Find Web sites on the Internet that contain helpful information for (1) parentsof children who stutter, (2) school-age children who stutter, and (3) adults who stutter.
5. One of the challenges for clinicians is to get a good speech sample from achild who may be somewhat shy or reluctant to talk to someone she doesn’t know well. Experiment with different ways of interacting with a child until you find a “best” method. For example, try asking lots of questions, try just playing quietly alongside a child, and try playing with a child and making comments about things you are playing with together.
Conture, E. (2001). Assessment and evaluation. In Stuttering: Its Nature, Diagnosis, and Treatment. Boston: Allyn & Bacon.
In this chapter, Conture covers many details of the assessment not dealt with in the chapter that you have just read. Among these are finer points of audio and video recording, general interview procedures, and analysis of the speech sample. Conture also discusses concomitant problems like attention deficit hyperactivity disorder, Tourette’s syndrome, neuromotor problems, and word finding problems.

Guitar, B. (2010). Stuttering. In M. Augustyn, B. Zuckerman, & E. Coronna (Eds.), The Zuckerman Parker Handbook of Developmental and Behavioral Pediatrics for Primary Care. Baltimore: Lippincott Williams & Wilkins.
This brief chapter for pediatricians summarizes key questions and important information for parents, criteria for referral, and initial treatment strategies.
Richels, C., & Conture, E. (2010). Indirect treatment of childhood stuttering: Diagnostic predictors of treatment outcome. In B. Guitar & R. McCauley (Eds.), Treatment of Stuttering: Established and Emerging Interventions.
Baltimore: Lippincott Williams & Wilkins.
This unique chapter uses the authors’ Communication-Emotion Model of Childhood Stuttering as a rationale for their thorough evaluation of a child’s speech and language, as well as emotional factors in the child’s life. Variables measured before treatment predict short- and long-term outcomes.
Shafir, R. Z. (2000). The zen of listening: Mindful communication in the age of distraction. Wheaton, IL: The Theosophical Publishing House.
This is an excellent introduction to the practice of careful listening. Shafir is a speech-language pathologist who has developed her ability to listen to clients and writes eloquently about the healing powers of mindful listening.
Yairi, E., & Ambrose, N. (2005). Assessment of early stuttering. In E. Yairi & N. Ambrose (Eds.), Early Childhood Stuttering. Austin, TX: Pro-Ed.
This chapter gives a thorough description, based on assessment experience with hundreds of children, of how to obtain samples and analyze speech of preschool children who stutter. Information regarding prognosis is given by the experts.
Yairi, E., & Ambrose, N. (2005). Parent involvement and counseling. In E.
Yairi & N. Ambrose (Eds.), Early Childhood Stuttering. Austin, TX: Pro-Ed.
The authors critically review the literature on ways in which parents can be involved in treatment of early childhood stuttering and conclude that there is little evidence to support any of the approaches.
Yaruss, S. (2002). Facing the challenge of treating stuttering in the schools: Part 1. Selecting goals and strategies for success. Seminars in Speech and Language, 23, 153–159.
This volume of “Seminars” is a rich source of information for school clinicians. Experienced clinicians, many of whom work in the schools, have written chapters on a wide variety of topics, including interpreting IDEA ’97, doing an evaluation in a school setting, and planning therapy for school-age children.
Preliminaries to Treatment

Clinician’s Attributes
A Preference for Evidence-Based Practice
A Commitment to Continuing Education
Critical Thinking and Creativity
Clinician’s Beliefs
Treatment Goals
Reduce the Frequency of Stuttering
Reduce the Abnormality of Stuttering
Reduce Negative Feelings about Stuttering and about Speaking
Reduce Negative Thoughts and Attitudes about Stuttering and about Speaking
Reduce Avoidance
Increase Overall Communication Abilities
Create an Environment that Facilitates Fluency
Therapy Procedures
Procedures to Help Clients and Families Deal with Their Emotions Associated with Stuttering
Procedures to Reduce the Frequency of Stuttering
Procedures to Reduce the Abnormality of Stuttering
Procedures to Reduce Negative Thoughts and Attitudes about Stuttering and Speaking
Procedures to Reduce Avoidance
Procedures to Increase Overall Communication Abilities
Procedures to Create an Environment that Facilitates Fluency
Motor Learning Principles for Treatment
After studying this chapter, readers should be able to:
• Understand the most important attributes of an effective stuttering clinician
• Understand how the clinician’s beliefs about the nature and development ofstuttering affect her choices about treatment procedures at different ages
• Understand what important goals are for stuttering therapy and how they mayvary depending on the client’s age
• Understand the therapy procedures used to meet the goals selected for a client’streatment
Therapy protocol: A detailed plan for carrying out treatment
Empathy: Ability to put oneself in another’s place and to identify with the feelings that the other person has
Warmth: A feeling of positive regard for another person, often conveyed by tone of voice or body language
Genuineness: Honesty about life, oneself, and other people; a sense that the individual is comfortable with herself and speaks in a straightforward manner in which actions are congruent with thoughts, beliefs, and attitudes
Evidence-based practice: A commitment to use research evidence, client goals, and clinician expertise when choosing assessment and treatment tools
Critical thinking: An attitude of mind that encourages questioning; in the current context, the questioning is about whether a treatment will be effective for a particular client
Clinician’s beliefs: The perspective a clinician takes on the nature and development of stuttering that leads to her choices of assessment and treatment strategies
Abnormality of stuttering: Characteristics of an individual’s stuttering that make it stand out or be distracting to the listener. Examples are facial tension and facial grimacing, body movements used to release the stutter, and avoidances that make it difficult for the listener to follow what the individual is saying such as “well, um, you see, that is…”
Fluency-facilitating environment: A climate in one or more situations that makes it easier for an individual who stutters to speak more fluently
Emotions related to stuttering: Physiological responses and conscious feelings that arise in a person who stutters, in regard to the act of stuttering or the fact of being someone who stutters
Cognitive behavior therapy (CBT): Treatment based on the notion that persons’ perceptions of and thoughts about situations and themselves determine their feelings and behavior. CBT aims to help the persons see situations and themselves more realistically and more compassionately
Before presenting the details of treatment in the next four chapters, I want to provide some background for the therapy protocols you can use, as well as how and why you use them. First, I’ll describe important attributes of the clinician who works with people who stutter and how his beliefs about the nature of stuttering influence treatment decisions. Then I’ll discuss commonly held goals for stuttering therapy and finally the procedures to meet them.
The clinician is probably the most important ingredient in stuttering therapy other than the client. A clinician’s knowledge, skills, and personality have a major influence on outcome. This is true whether therapy’s major focus is to change behaviors, thoughts, or feelings–or some combination of these. In this section, I discuss some of the attributes that I think make a clinician effective, and I suggest how these can be developed. Unfortunately, there are no data that I’m aware of to support the importance of these attributes for stuttering therapy. In the field of psychotherapy, Carl Rogers (1961) and others have spent a considerable amount of time and effort measuring the effects of some of these clinician attributes on treatment success. In the field of stuttering treatment, Manning (2010) and Zebrowski (2007) have excellent chapters on the role of the clinician and the importance of the clinical relationship.
Much of my thinking about the treatment process has been influenced by my experiences as a client and later as a student of Charles Van Riper. He was a master clinician of stuttering therapy. Let us begin then with Van Riper’s (1975a) description of three important clinician characteristics: empathy, warmth, and genuineness.
Empathy, in this context, is the ability to understand the feelings, thoughts, and behaviors of someone who stutters. You might think that this is a little easier for clinicians who stutter. However, Van Riper’s own clinician, Bryng Bryngleson, was a fluent speaker who showed an impressive understanding of individuals who stutter. Once, Bryngleson assigned Van Riper the task of voluntarily stuttering to 10 strangers, but he was unable to carry it out. Exhausted from trying again and again and failing over and over, he sought out Bryngleson in his office. Bryng, as he was called, jumped up from his chair and headed for the door, saying, “It’s OK, Van, just follow me and watch.” Bryng then went into a nearby tobacco store, walked up to the clerk, and pretended to stutter–with the longest, loudest stutter that Van Riper had ever heard, causing the clerk to cower behind the counter. Van Riper was astounded. That single demonstration by his clinician had a huge impact on Van Riper. He felt deeply supported by Bryngleson’s acceptance of his failure and Bryngleson’s willingness to risk ridicule to help him. Remember this when you wonder how to show your empathy.
You can also develop empathy with all your clients by working on your ability to listen deeply and acceptingly. It will help to observe body language, posture, and the words they use. Van Riper said that he could improve his understanding of a client’s feelings if he assumed the same body posture that the client had. You can also get some idea of what clients experience by going out in public like Bryngleson did and stuttering voluntarily, though you don’t have to stutter as long or as loud as Bryng did. Reading stories written by people who stutter and parents of children who stutter will also help you better understand the experiences that have shaped their feelings. Good examples of such writings are Living With Stuttering (St. Louis, 2001) and Forty Years after Therapy: One Man’s Story (Helliesen, 2002). I describe these books and others in the suggested readings at the end of this chapter.
This attribute has also been referred to as “unconditional positive regard” (Rogers, 1957). Much of it is conveyed in the tone of voice, facial expression, and body language of the clinician. Clients whose clinicians demonstrate warmth feel accepted, liked, and nurtured. Warmth creates an environment that supports learning and helps clients make difficult changes. Warmth is also expressed in the comments the clinician makes when a client has done something well. This is sometimes harder than you think. It surprises me when I watch a videotape of one of my therapy sessions, how many opportunities I miss reinforcing the client with a “Good!” or “Well done!” Therefore, I try to watch videos of myself working with a client to discover things that I need to work on. Although it is initially painful for all of us to watch video recordings of ourselves, this is one of the best ways we can improve. Clinicians should become aware of how much or how little enthusiasm they show and warm encouragement they give to their clients. These are important tools of therapy.
A third characteristic of good clinicians that Van Riper (1975a) described is “genuineness,” which he equated with Rogers’s (1961) “congruence.” Both terms refer to a clinician’s honesty and self-acceptance. She just tries to be who she is, “roughness, pimples, warts, and everything” as Oliver Cromwell said on having his portrait painted. Genuineness allows clinicians to be honest with their clients, not sugarcoating the hard lumps of reality that must be swallowed if real progress is to be made.
For example, Van Riper said to one of his clients with his characteristic bluntness, “Why do you have to have all that junk in your speech? Can’t you just go ahead and say the word, starting with the first sound and working your way through it slowly, syllable by syllable?” (Van Riper, 1975b). When a client senses the clinician’s genuineness, he gains trust and begins to believe that his clinician means it when she asks about his thoughts and feelings that he can let go and honestly express his frustration, fear, hate, and anger, convinced that the clinician will understand and accept him and his feelings, and be strong enough to be unhurt by them. Clinicians can cultivate their genuineness and strength by being open about their limitations and learning self-acceptance through psychotherapy, spiritual practice, or other experiences that help them accept both their weaknesses and their strengths.
A Preference for Evidence-Based Practice
There are more traits than those three listed above that characterize good clinicians. One is a clinician’s desire and ability to base her clinical practice on evidence of its effectiveness, called evidence-based practice. In choosing tools and approaches for evaluating and treating someone who stutters, a clinician who wishes to grow looks for evidence of the effectiveness and appropriateness of treatment approaches she is using or considering. She works together with the client or family in the diagnostic evaluation to determine which treatment approach is likely to meet the client’s or family’s goals most effectively. She measures the client’s progress during treatment to assess whether this approach is working with this person. She is flexible, creative, and insightful enough to find ways of altering treatment if it is not working. Many treatments, including some described in this text, have relatively little published data that support their effectiveness. For example, my approaches to school-age children and adolescents/adults have been derived from years of experimenting, and only now have I developed enough consistency to start collecting data to be disseminated. This does not preclude their being used, but a clinician should be careful to assess how well they work for her clients with measures made before, during, and after treatment. In addition, for those clinicians who find them very effective, obtaining research evidence on them to share with others would represent a valuable contribution to the field. Ideas and information on evidence-based practice can be found in Bothe (2004), Frattali (1998), Guitar (2004), Guitar and McCauley (2010), Piertranton (2003), and Sackett, Straus, Richardson, Rosenberg, and Haynes (2000). The American SpeechLanguage-Hearing Association (ASHA) has provided member access to useful tutorials on evidence-based practice at
An interesting example of an early attempt at evidence-based practice is the data that Van Riper (1958) kept as he experimented with different forms of treatment for stuttering. He attempted to write down in detail what variations he made with his treatment protocols each year and reassessed his clients five years after they had finished therapy. Although he admits his methods are not perfect, the chapter in which he presents 20 years of experiments in stuttering treatment is a fine example of evidence-based clinical practice for its time, some 55 years ago. In this spirit, in the treatment chapters that follow, I suggest ways in which clinicians today can measure client progress.
A Commitment to Continuing Education
Another important attribute for clinicians is the habit of continually updating knowledge gained in graduate school. New methods of evaluation and treatment are developed every year, and new data on treatment effectiveness become available. It is vital for the clinician to keep up to date with the latest and best practices. Journals are the best source of this information, but recent editions of books that review diagnostic and treatment methods for stuttering can also be helpful. New approaches to treatment often require training. Short courses at the annual ASHA convention and workshops offered through schools, hospitals, state associations, and other institutions are excellent sources of such training. However, before adopting a new approach, a clinician should critically analyze the quality of evidence that supports its claim to effectiveness.
Critical Thinking and Creativity
Clinicians should become discriminating consumers and ask, “Which new diagnostic tools and treatment approaches are effective, and which clients are they appropriate for?” This demonstrates critical thinking. Some new approaches are not all they are cracked up to be. For example, many years ago, a well-known psychologist and his colleagues (Azrin & Nunn, 1974) suggested that teaching clients simply to take a breath and relax before speaking was an effective treatment for stuttering. Researchers at another clinic tested the approach and found it to be far less effective in their clinic than its developers had suggested (Andrews & Tanner, 1982). Nevertheless, there may be some aspects of relaxation and breathing that are useful for some clients in the hands of a clinician who becomes skilled at integrating these tools into a broader approach.
Another critical question is “Will this approach work for my clients in my environment?” Often a treatment that works under laboratory conditions with carefully selected subjects does not work as well in the real world of a public school, for example. But a clinician may be able to adapt an approach to suit her situation. For instance, an approach developed for very young children in tightly controlled clinical studies with total fluency as its goal may need to be altered so that some degree of easy and open stuttering is an acceptable outcome when used with older children.
It is important for clinicians to weigh their beliefs about the nature of stuttering against the available data, then develop clinical procedures compatible with the clinician’s beliefs–procedures supported by data, ideally data collected by the clinician as well as others. My own beliefs about the etiology and development of stuttering that were presented in the first few chapters are reviewed here only in enough detail to illustrate the relationship between beliefs and treatment procedures. I believe that predisposing physiological factors interact with developmental and environmental influences to produce or exacerbate core behaviors that often (but not always) begin as repetitions. When some children experience these early disfluencies as frustrating and/or embarrassing, they increase the amount of physical tension and speed they use when experiencing disfluency. This then creates for these children secondary or coping behaviors as well as negative feelings and attitudes. They learn escape behaviors through instrumental conditioning, speech fears through classical conditioning, and word and situation avoidances through avoidance conditioning. All of these factors and how they contribute to stuttering are reflected in the stages of development I described in Chapter 7.
How does this point of view about the etiology and development of stuttering affect treatment? Let’s use the management of school-age children who stutter to illustrate this point. In my view, a child’s treatment plan is determined by his developmental level of stuttering, and each advance in level requires new components in treatment. A first grader with beginning stuttering who is not embarrassed or afraid to talk and who doesn’t avoid talking requires a different treatment than a fifth grader with intermediate-level stuttering who has developed fears and avoidances in response to his stuttering. In my view, the first grader may be treated with an approach that focuses on increasing fluency and deals only minimally with negative feelings and avoidance behaviors. On the other hand, the fifth grader needs help to reduce the tension and struggle and the fears and avoidances. In contrast to these ideas about treatment, a clinician who doesn’t believe that fears and avoidances are crucial in understanding and treatment of stuttering might treat both children with the same approach.
Another way in which a clinician’s beliefs can affect management is in the assessment procedures she uses. Assessment tools should provide clinicians with information that is essential for planning treatment and measuring progress. In evaluating the children described above, I would evaluate each child’s feelings and attitudes about his speech, as well as his use of word and situation avoidances, to accurately determine each child’s developmental/treatment level and decide which aspects of the problem to focus on first. Another clinician–for example one who is atheoretical or unconcerned about the etiology and development of stuttering–might simply want to measure each child’s frequency and severity of stuttering.
A third way in which a clinician’s beliefs about the nature, development, and treatment of stuttering can affect clinical behavior relates to how she counsels the parents of her clients. In counseling the parents of these two school-age children, my beliefs would guide me to describe the etiology of stuttering as being unknown at present, but probably related to the way a child’s brain processes speech and language. Using terminology appropriate to the parents, I would talk about brain processing that may predispose a child to stutter, and I would emphasize that this suggests that parents don’t cause stuttering. I would also explain that the child’s way of processing speech and language can be changed, which means that parents can be vital in helping a child overcome or manage his stuttering. I would also discuss with parents the importance of factors in the environment that might be contributing to the child’s stuttering problem and discuss ways of modifying these factors. Lastly, I would use my understanding of the development and nature of stuttering to give parents a general idea of the course of therapy and possible outcome. Clinicians with other beliefs might not go into the nature of stuttering because they feel it is not well-understood and would instead just counsel the parents about their role in the child’s treatment.
Treatment goals will vary with a clinician’s beliefs, the client’s age, and the developmental/treatment level of his stuttering. Nonetheless, it is still possible to describe most of the goals that clinicians have for clients who stutter and to suggest which goals are likely to be paramount for which level. Individuals differ in their strengths and weaknesses at the outset of treatment, and these change as treatment proceeds. Thus, a clinician needs to ask herself: “What does this client need? What does he need from me? What does he need from me right now? And why?” (Van Riper, 1975a, p. 477).
As I have said earlier, the clinician is not the only one who determines the goals of treatment. Clients and their families have an important role to play in choosing goals that are paramount for them. Ongoing discussions between clinicians and clients about treatment goals strengthen clients’ motivation to achieve them and enhance the relationship between clients and clinicians. The following statement by Donald Baer (1990), an eminent behavioral psychologist, expresses this philosophy.
“It seems only reasonable to learn that when stutterers are given control of the therapeutic consequences that presumably can change their output, some of them choose different targets than would their therapists or, probably, other stutterers, and some of them target not so much their speech output as they do a private response that they describe as sense of ‘imminent loss of control. ’” (p. 35)
The selection of goals presented in this section owes much to the Guidelines for Practice in Stuttering Treatment (American Speech-Language-Hearing Association, 1995)
Reduce the Frequency of Stuttering
This can be achieved in a variety of ways, but it is important to reduce the frequency of stuttering without creating other behaviors, such as taking deep breaths before speaking that may be distracting to the listener (and speaker) and may therefore hamper communication. This goal is appropriate for all ages and levels of stuttering; note that for preschool children, the goal should be to reduce frequency of stuttering to essentially zero.
Reduce the Abnormality of Stuttering
I think much of the abnormality of stuttering comes from the conditioned tension and struggle behaviors that occur during moments of stuttering. It shows up as squeezing of facial muscles as the person is trying to say a word that is blocked. Reducing this tension and struggle is an important goal for school-age, adolescent, and adult clients. In addition, behaviors that occur before the stutter (avoidance) and behaviors that are deployed to terminate the stutter (escape) should be eliminated or at least greatly diminished. These include (1) avoidance behaviors such as the repetition of the sound “uh” before saying a word, and (2) escape behaviors such as eye blinks and head nods used to terminate a block. For some school-age children and older clients, it may not be possible to eliminate their stuttering. Instead, the stuttering can be changed so that it is easy and comfortable both for the speaker and the listener and doesn’t interfere with communication. Van Riper and other experienced stuttering clinicians have suggested that a stutterer may not always have a choice whether he stutters but he does have a choice about how he stutters. This choice includes stuttering in a way that is easier and briefer than his old habitual pattern. This new way of stuttering reduces fear because it feels and sounds more like normal speech and is often unnoticed by the listener. Once the person has confidence in his ability to stutter this easily, he is less likely to increase muscle tension in response to an actual or anticipated stutter.
Reduce Negative Feelings about Stuttering and about Speaking
Many individuals who stutter appear to have a temperament that is sensitive and somewhat perfectionistic. They are thus vulnerable to feelings of embarrassment, fear, shame, and other negative feelings associated with their stuttering. A cycle can develop in which stuttering gives rise to negative feelings, which in turn increase tension and other struggle behaviors, which then generate more negative feelings. Classical conditioning plays a major role in this cycle. Therefore treatment strategies to deal with classically conditioned behaviors, such as deconditioning and counterconditioning, are crucial in treatment. These and related strategies are discussed in the chapters that describe treatment of school-age children, adolescents, and adults. Reducing negative feelings is an important goal for many clients beyond age 6 or 7, although a few school-age children and even older clients may not have strong negative feelings about being someone who stutters. They may, however, have feelings of frustration about their impediment to speaking easily. Most of these feelings can be changed significantly–either directly or indirectly–through treatments that give the client repeated experiences with effective communication and ease of speaking. One major difference among treatment approaches is whether they deal with clients’ negative feelings and emotions directly or indirectly.
Reduce Negative Thoughts and Attitudes about Stuttering and about Speaking
In the chapter on the development of stuttering, I described how people who stutter may acquire negative self-concepts through repeated experiences of stuttering and perceiving–sometimes correctly and sometimes incorrectly–that listeners are impatient or disapproving. As these perceptions become more and more deeply ingrained, they begin to affect a stutterer’s expectations in speaking situations. This can lead to more stuttering. If a stutterer expects rejection or disapproval, he may try very hard not to let the stuttering out by adopting fixed, tense articulatory postures that trigger blockages. These are often devastating to the individual who stutters and make him feel helpless and out of control. This then “snowballs” downhill, gathering speed, from negative thoughts to more stuttering to more negative thoughts, on and on. This avalanche of events is at the heart of much chronic stuttering.
Good treatment can roll the snowball back uphill. Clients can be toughened up (desensitized) to the experience of stuttering, decreasing their fears and negative expectations. They can also be shown how to say their feared words without as much struggle. As a result, they will approach speaking opportunities with more relaxed speech muscles and find themselves stuttering more easily or not stuttering at all. This in turn will lead to more positive expectations, which can lead to easier or less stuttering; thus a positive cycle begins to replace the negative one that preceded it.
A different approach is to give the client repeated experiences of being fluent–in many situations, with increasing linguistic and social demands over a relatively long period of time and with much success and little failure. The aim is to replace expectations of stuttering with expectations of fluency. Somehow, through habit change accompanying changes in how the brain produces speech, good quality normal fluency can be established and maintained. This seems to work best, in my experience, with younger clients and when used with adults, with those individuals who have few avoidance behaviors. For those with much avoidance, the strategies described below can help.
Reduce Avoidance
Avoidance behaviors, as you will remember, are evasive maneuvers taken by individuals to keep from stuttering. Sometimes they may occur very close in time to the expected stutter, such as saying “um” or “well” just before attempting to say a feared word. Other times they may be quite separated in time from the expected stuttering, such as not volunteering to be in a school play or by driving 20 miles to talk to someone rather than telephoning her. Some stutterers may have an innate predisposition to avoid because of their temperaments, as described in Chapters 2 to 6. Avoidances keep stuttering “hot,” because they prevent an individual from learning that it is possible to stutter in an easy fashion and communicate well. Reducing avoidance is usually not the first treatment goal on the list, although it may be one of the most important goals for more advanced levels of stuttering. Usually, before helping clients reduce avoidances, clinicians need to help them reduce negative emotions about stuttering and teach them to stutter more easily. Reducing avoidances is a major goal for older children and adults, but again, some approaches work indirectly by giving them tools to increase fluency, which then, one hopes, reduces fear and thus decreases avoidances.
Increase Overall Communication Abilities
The ability to communicate easily and well varies a great deal from client to client. It may be affected by severity of stuttering, temperament, avoidances, and communication models in the family. For many of us who work with individuals who stutter, effective communication is an important treatment goal. Some clients will become good communicators once the frequency and severity of their stuttering, along with their negative feelings and attitudes about speaking and stuttering, have been reduced. For other clients, guided practice and structured experiences in communication are essential. Once clients feel they can communicate easily, they often begin to seek out talking experiences, their avoidances drop away, and they become comfortable and open about any remaining stuttering that occurs. The goal of effective communication is most needed for older children and adults who have developed avoidances. Many of these clients, especially those with more severe stuttering, have been preoccupied with their stuttering and not spent much time learning to communicate effectively (Curlee, personal communication, March 3, 2004). They may still have hesitancies and a herky-jerky style of speaking that lacks the fluidity of normally fluent speech.
Create an Environment that Facilitates Fluency
This goal is paramount for working with young children who can often be treated by helping the family reduce pressures on the child’s speech and increase positive aspects of the child’s speaking environment to create a fluency-facilitating environment. For example, family members can spend one-on-one time with the child, using a slow speech rate and careful listening skills, thereby increasing the child’s daily opportunities to experience fluency. The child’s environment may be made more positive through praise and appreciation of his fluent speech and/or his other accomplishments. This goal of improving the child’s speaking environment may also be important for school-age children. However, teachers and aides, as well as family members, need to be enlisted in facilitating the child’s fluency. Older clients can make their environments facilitating to both fluency and easy stuttering by being open about their stuttering and sharing with others how listeners can be most helpful to them.
The aim of this section is to outline the tools and strategies that clinicians can use to work on each of the treatment goals described above. By understanding which procedures are most likely to be useful in achieving each goal, clinicians can select those procedures that best suit each client and are in accord with their own beliefs. The procedures outlined here are fully described in the therapy sections on each developmental/treatment level.
My belief about stuttering treatment is that the emotions associated with stuttering must be understood and dealt with if therapy is to be successful. Therefore, I begin discussing procedures with a substantial discussion of how the clinician is involved in this process.
Procedures to Help Clients and Families Deal with Their Emotions Associated with Stuttering
In many books, this section might be called “counseling” but I prefer to describe this process as working with emotions related to stuttering. The clinician’s attributes of empathy, warmth, and genuineness are vital to dealing with the emotions that can impede changes in stuttering and to fostering the emotional processes that can foster recovery. Although there are similarities in how the clinician responds to the parents of a child who stutters and how she helps older children and adults, there are enough differences that I describe them separately.
In working with preschoolers, working with the child’s family is paramount. The family must develop faith not only in the clinician’s abilities but also in her understanding and acceptance of their feelings. In the initial meetings, the family is likely to feel anxious because they have watched their child struggle, for weeks or even months, not knowing how to help him. The clinician must listen with care to how the parents (or other family members) describe the child’s stuttering, his response to it, and their own responses to it. Listening to their feelings throughout the course of treatment will provide support for the family and will convey the all-important idea that they and the clinician form a team working together to help the child. As the clinician listens, she acknowledges their feelings, sometimes restating them to make sure she understands. She refrains from simple reassurance, but instead, when appropriate, shows them evidence that the child is making progress, shares information about recovery, or helps them realize all the positive things the child has going for him. This should be done, not instead of, but in addition to, acknowledgment of feelings they may express–sometimes of frustration, discouragement, and guilt, and sometimes of hope and pride. In summary, in working with young children and their families, the clinician must be aware that the family’s emotions can be a part of the reason that stuttering may worsen after onset and that treatment can reverse this.
In treating clients older than preschoolers, the clinician works more with the client directly, rather than his family. The clinician’s role–in helping the client deal with emotions–is to create an atmosphere in which the client feels more and more comfortable expressing feelings. These feelings may include shame and fear, and sometimes anger toward the stuttering, the clinician, and listeners. Such feelings should be accepted as normal emotions that occur when a person has stuttered for a number of years and has experienced listeners’ impatience, rejection, or even teasing.
In the early stages of therapy, as the client learns more about his own stuttering, the clinician’s genuine interest and curiosity about the stuttering have the potential to counteract the negative feelings the client may have–expressed or unexpressed– toward his stuttering. The goal is for the client to accept his stuttering and take responsibility for changing it. For this to happen, the clinician must show her acceptance of it. Another goal is to help the client build up tolerance to the frustration and the fear of being stuck. This must be achieved in order to reduce tension and struggle. Activities to accomplish this include stuttering openly, without avoidances, and stuttering voluntarily on nonfeared words. These activities usually cause an increase in emotion, giving the clinician an opportunity to help the client explore and accept his feelings. The clinician simply asks the client how he’s feeling about doing these tasks, listens attentively, and doesn’t ask the client to work on his stuttering as he talks about feelings.
The next goal is for the client to learn to stop rewarding the old struggle behaviors. This is done by progressively changing the form of stuttering from tense and rapid to loose and slow. To help the client do this, some clinicians teach what are called cancellations, pull-outs, and preparatory sets, which are explained more fully later. As this goal is being worked on, many emotions and resistances usually arise, and these are opportunities for the clinician to help the client express them, explore them, and accept them. Previously, the client may have been afraid of listeners’ reactions to his stuttering; now he may be afraid of listeners’ reactions (real or imagined) to his managing his speech by slowing his speech rate, using easy onsets of phonation, or prolonging stutters beyond the moment when they can be released. These are things that the client may have to do to reduce tension and struggle, but they are often hard to do with friends, family, or strangers. The clinician can help the client learn that by and large, listeners respect a person who is openly working on his challenges. And when listeners are impatient, anxious, or rejecting, it is because they have their own issues, and the client can learn to tolerate and transcend these reactions.
There is a rhythm to treatment of older children and adults who stutter. The clinician guides the client to carry out various tasks. Before, during, or after carrying them out, the client feels various emotions that are related to the stuttering and to changing it. The clinician helps the client express these emotions, and together they accept them, then move on to work on the next step together. The clinician’s role in this process has been described in this way:
“(People who stutter) need a permissive figure to whom they can ventilate their anxieties and frustrations. They need a companion who can share their difficulties in communication without becoming punitive or upset. They need someone to point out a possible pathway out of communicative deviancy and who will stay with them even when they fail. They will learn all they need to about themselves by working with their stuttering. Perhaps what we are saying is that the stutterer needs a very good teacher rather than a psychiatrist.” (Van Riper, 1958, p. 381)
Procedures to Reduce the Frequency of Stuttering
Operant conditioning procedures are often part of treatment approaches for achieving this goal and typically involve reinforcement for fluency and mild punishment for stuttering. Rewards may be verbal, such as the clinician’s praise or approval, or tangible, such as tokens that can be redeemed for snacks, prizes, or an opportunity to take a turn in a game. Mild punishment may simply be calling attention to a stutter or requesting the individual to try the word again. Rewards for fluency and mild punishment for stuttering can be the primary tools used for beginning stuttering and are often coupled with a hierarchy based on the complexity and length of utterances. In this case, clients move from producing one or two words fluently through longer phrases to spontaneous speech. Reward and punishment may also be used as “shaping” tools for intermediate or advanced-level stuttering, in which clients begin by speaking in a way that produces instant fluency, such as speaking very slowly, and then progressing to more and more normal-sounding speech in more and more difficult situations. This approach–sometimes called “prolonged speech”–was foreshadowed in remarks by Francis Bacon in the late 1700s (Siegel, 2007). Here are Bacon’s words quoted by James Boswell:
“In all kinds of speech, either pleasant, grave, severe, or ordinary, it is convenient to speak leisurely, and rather drawlingly than hastily: because hasty speech confounds the memory, and oftentimes, besides the unseemliness, drives a man either to stammering, a non-plus, or harping on that which should follow; whereas a slow speech confirmeth the memory, addeth a conceit of wisdom to the hearers, besides a seemliness of speech and confidence.” (Boswell, 1791)
The general term for treatments such as prolonged speech that focus on increasing fluency rather than decreasing the abnormality of stuttering is “fluency shaping.”
Procedures to Reduce the Abnormality of Stuttering
These procedures are appropriate for clients who have developed struggle, tension, escape, and avoidance behaviors that make their stuttering stand out as abnormal to the listener and the client himself. Therapies that target the abnormality of stuttering often use reward and mild punishment to change long, tense stutters into increasingly briefer and more relaxed ones and to diminish clients’ use of escape and avoidance behaviors. To meet this goal, reward and punishment are often accompanied by a systematic program for reducing negative emotions. Such programs are founded on the belief that negative emotions elicit increased tension, escape, and avoidance behaviors and that these behaviors are maintained by the fact that they are rewarded when the stutterer finally gets the word out by squeezing and pushing on it. These approaches are often referred to as “stuttering modification.”
A classic stuttering modification approach is that of Van Riper (1958, 1973, 1975b), which begins first by reducing negative emotions through objective study of the stuttering, then focuses on desensitization to the frustration and embarrassment of it. Next, the clinician teaches the client to self-correct his stuttering after a stutter, then during a stutter, then before the stutter occurs. Stuttering modification and fluency shaping often result in a modified style of speaking, which contains brief disfluencies which are produced in a slightly slower than normal way of talking.
Some therapy approaches–both fluency-shaping approaches for older children and adults as well as therapy approaches for preschoolers–don’t aim to reduce the abnormality of the stuttering behavior directly, but instead focus on increasing fluency with the assumption that as fluency increases, stuttering diminishes to negligible levels.
Procedures to Reduce Negative Thoughts and Attitudes about Stuttering and Speaking
There are a number of therapy procedures that can help clients become more realistic about how listeners perceive them and what this may mean to them. Cognitive therapy, for example, can be an excellent technique for helping clients think and feel more positively about their speech, listeners, and the situations that have elicited negative emotions in the past. Clients can learn to examine their thought processes and understand how what they think influences what they feel and how they act, particularly in regard to such maladaptive behavior as muscular tensing that leads to more stuttering. Some clinicians use cognitive behavior therapy as their sole treatment and others as a supplement to techniques for learning to speak fluently or to stutter in an easier way. The book, Cognitive Therapy: Basics and Beyond (Beck, 1995), is a good source for learning this approach, and I discuss cognitive therapy in the chapter on advanced stuttering. An excellent introduction to this approach with stuttering are two Stuttering Foundation DVDs entitled Tools for Success: A Cognitive Behavior Therapy Taster and Implementing Cognitive Behavior Therapy with SchoolAge Children (
Procedures to Reduce Avoidance
Some clients have very little avoidance, and once they learn to speak fluently, they enter speaking situations freely without expectation of difficulty. Others, however, because of temperament, learning, or both, have a strong tendency for avoidance. Only minor avoidances may appear in beginning stuttering, but it is a problem that must almost certainly be addressed at the intermediate and advanced levels. Treatment to reduce avoidance should begin by reducing negative emotions, particularly fears of stuttering and of listeners’ reactions. Fear of stuttering can be tackled by rewarding clients with praise, support, or tangible reinforcement for “catching” a stutter and holding onto it. Fear of listeners’ reactions can be lessened by clients’ voluntarily stuttering to acquaintances and strangers. When a stutterer can deliberately imitate his typical stuttering pattern and pretend to stutter, he finally feels in control during a stutter; this and the feeling of “stuttering” while also feeling in control is highly rewarding.
Reducing fear is not enough, however. Studies of animal behavior have shown that, even when avoidance symptoms disappear after fear is reduced, fear eventually returns and so do its symptoms–conditioned avoidance behaviors (Ayres, 1998). Thus, new responses to the old stimuli must be taught. In stuttering therapy, an example of learning a new response to an old stimulus is for a stutterer to slow his speech rate as he says a word he expects to stutter on. This is an aspect of the “preparatory set” used in many stuttering modification approaches, as well as the “downshifting” to a slower rate before attempting a difficult word, taught in fluency-shaping programs.
Avoidances are not confined to the moment just before a difficult word. Individuals who stutter may also avoid opportunities to speak by pretending to be busy when the telephone rings or by waiting for someone else to make introductions of new acquaintances. These avoidances can be treated by helping a client construct a hierarchy of easy-to-difficult speaking situations, in which he can use newly learned stuttering modification or fluency-shaping techniques. Clinicians can also motivate clients to continue seeking out new situations in which they can be open about their stuttering and can use their new strategies to manage stuttering. At meetings of the SpeakEasy Associations of Australia and the United States and conventions of the National Stuttering Association, there are always impressive testimonials by clients who have sought out public speaking opportunities, joined Toastmasters (an international organization of people who want to practice public speaking), or found other ways of increasing their approach behaviors and decreasing their tendency to avoid stuttering and speaking.
Procedures to Increase Overall Communication Abilities
For many children, adolescents, and adults, communication blossoms when fears of stuttering and listeners’ reactions are reduced, and ease of speaking is increased. For others, longstanding habits of avoiding speaking situations and the accompanying lack of social experience have stunted the growth of their communication skills. For still others, concomitant problems, such as attention deficit or extreme shyness, may have prevented them from learning how to communicate well. Communication skills should be addressed in treatment whenever it appears that they are not appropriately developed. Observations of a client’s communication and reports from a school-age child’s teachers will indicate the areas that may need to be addressed. Specific skills that can be worked on include eye contact, turn taking, maintaining a topic, making relevant contributions to conversation, speaking intelligibly, clarifying and repairing what was said, and developing a willingness to initiate and maintain communicative interactions with others (Kent, 1993; Smith, McCauley, & Guitar, 2000). Although these skills can be worked on individually, group therapy provides excellent opportunities for clients to practice them. Direct instruction, modeling, role-playing, and video-recorded feedback with discussion can be used to teach and refine communication skills.
Procedures to Create an Environment that Facilitates Fluency
Preschool-age children, especially those on the borderline between normal disfluency and stuttering, may need only a little change in their environments for their stuttering to disappear permanently (Starkweather, Gottwald, & Halfond, 1990). Treatment focuses on parents: counseling them to reduce their anxieties, modeling for them, and continuing to support the changes they make. Parent-child interactions are usually the key element of the environment that can be changed to facilitate fluency. Video recordings and playback of these interactions in the clinic or observations at home, coupled with parent counseling, can help parents improve how they communicate with their child (Guitar, Kopff-Schaefer, Donahue-Kilburg, & Bond, 1992). Parents usually work on creating a facilitating environment during brief, one-on-one daily sessions with the child. In some families, other aspects of the environment may need to be changed, such as the home’s hurried pace of life, stressful life events, and the communication styles of other family members. Older preschoolers may also benefit from a direct approach, involving contingencies for fluent speech and stuttering.
For school-age children, the creation of facilitating environments may include working with the child’s family, but the school setting may be equally important, if not more so. Clinicians often work in partnership with a child or adolescent to make school a “fluency-friendly” environment. The clinician may arrange meetings with the child and his teachers to improve the teachers’ understanding of his stuttering and to open lines of communication between the child and teachers. A child’s peers can be invited to treatment so that they may improve their understanding of the child’s stuttering, while the process of the child’s openness about his stuttering with other children is begun. Freely discussing his stuttering with other students is one of the most powerful ways for a school-age child to make his environment more fluencyfriendly. For some children, a powerful boost can be given to therapy’s progress if they are able, with the clinician’s help and support, to make a presentation to their class about the nature of stuttering in general and their stuttering in particular.
Openness about stuttering is also a major way in which adults can create a supportive environment. By commenting on their stuttering, by showing a sense of humor about it, and by sharing what techniques they’re working on, adults who stutter can create environments in which their listeners are quite comfortable with the adults’ stuttering. This helps them feel free to use various fluency-enhancing techniques.
My own treatment for each age and the treatments of several other clinicians are presented in the next four chapters. Descriptions of the other clinicians’ approaches include their beliefs about the nature and development of stuttering as well as rationales for their choice of goals and procedures to treat each level. Both my own approaches and those of other clinicians have been developed and refined, usually over several years’ of trial and error. When possible, supporting data are provided for each treatment, but in many cases, where such data are not available, I suggest what data would be appropriate to gather.
Motor Learning Principles for Treatment
Most of our treatments involve working with the client in a relatively structured environment for a short period of time and then hoping for generalization to an unstructured environment between treatment sessions. The motor learning literature– cogently summarized by Verdolini and Lee (2004)–suggests that certain principles must be followed for generalization to take place. Table 10.1 summarizes some of their ideas and suggests how they might apply to the treatment of stuttering.
Table 10.1 Principles of Motor Learning and Their Applications to the Treatment of Stuttering

• The clinician’s attributes are a vital ingredient in treatment success.
• Empathy, genuineness, and warmth are three clinician attributes that have been considered important by Van Riper (1975a).
• An important component of best clinical practice is choosing evaluation proceduresand tools that have been shown to be valid and reliable.
• Best clinical practice dictates becoming aware of evidence of effectiveness for treatment procedures that you use and adapting treatment procedures to fit clients’ needs, as well as continuous assessment of improvement in attributes that have been chosen as goals for treatment.
• Continuing education is vital to keep abreast of new approaches and new evidenceof effectiveness of current approaches.
• It is important for clinicians to develop an informed set of beliefs about the nature ofstuttering and to fit assessment and treatment procedures to those beliefs.
• Goals for treatment and for continuing assessment should come from not only theclinician’s beliefs but also from the client’s (or family’s) informed choices.
• Treatment procedures for meeting these goals can include methods of reducing frequency and severity of stuttering and secondary behaviors, reducing negative emotions and thoughts that interfere with fluency, increasing communication abilities, and developing environments which facilitate fluency.

1. What are the three important characteristics of a clinician described by VanRiper?
2. How might each of these characteristics facilitate progress in treatment?
3. What are the characteristics of evidence-based practice?
4. How might two clinicians’ beliefs about the nature of stuttering result in two verydifferent treatment approaches? How might these beliefs result in two similar treatment approaches?
5. Which of the treatment goals described in this chapter are appropriate for borderline stuttering?
6. Which are appropriate for beginning stuttering?
7. Which are appropriate for intermediate stuttering?
8. Which are appropriate for advanced stuttering?
9. Describe the differences between “fluency-shaping” and “stutteringmodification” approaches to treatment.
10. How might reducing negative emotion reduce stuttering frequency?
11. How might reducing stuttering frequency reduce negative emotion?
12. Which goal would you start with and why?
1. Video record yourself and a client during an evaluation or a treatment session.The first time you watch it, note only the things you think you do well. The second time you watch it, note two things you would like to improve. Meet with a colleague or supervisor and discuss how to improve the things you would like to and then work on those in another session and videotape yourself again. Watch this new tape for improvements in the behavior(s) you have chosen to work on.
2. Choose a test you use in your evaluation procedures, and try to find evidenceof its validity and reliability.
3. Choose a treatment procedure you use, and search the literature to see if you
can locate any information about its effectiveness.
4. Find a stuttering treatment approach that is described in detail and determinewhat the goals of treatment are, what the procedures are to reach these goals, and whether there is a description of how to measure progress on these goals. Examples of such approaches are: Systematic Fluency Training for Young Children by Richard Shine (Pro-Ed, Austin, TX); Fun with Fluency-Direct Therapy with the Young Child by Patty Walton and Mary Wallace (Imaginart International, Inc, Bisbee, AZ); A Primer for Stuttering Therapy by Howard Schwartz (Allyn and Bacon, Boston); and Dynamic Stuttering Therapy by Barbara Dahm (
5. Describe in detail your own beliefs about the nature of stuttering applied tochildren with intermediate stuttering. Given these beliefs, what therapy goals do you have for a child with intermediate stuttering?
Bothe, A., (Ed.). (2004). Evidence-based treatment of stuttering: Empirical bases and clinical applications. Mahwah, NJ: Lawrence Erlbaum.
This text, available in hard copy or electronic format, contains chapters dealing with data on stuttering treatments and the scientific basis of treatment approaches.
Cordes, A., & Ingham, R., (Eds.). (1998). Treatment efficacy for stuttering: A search for empirical bases. San Diego, CA: Singular.
This is an excellent volume of papers by clinician-researchers who are searching for a scientific foundation for the treatment of stuttering. As the introduction makes clear, this book is the outcome of a continuing series of conferences on this topic.
Guitar, B., & Peters, T. (2003). Stuttering: An integration of contemporary therapies. Memphis, TN: Stuttering Foundation. (
This booklet describes in detail the two approaches mentioned in this chapter:
fluency shaping and stuttering modification.
Manning, W. (2010). Clinical decision making in fluency disorders (3rd ed.). San Diego: Singular.
The first chapter describes many aspects of the clinician as well as of the clinical interaction in stuttering therapy. Also relevant to the material discussed earlier is Chapter 6: Facilitating the Therapeutic Process. Manning describes goals for treatment and subtleties of how and when to work toward them. Chapter 7: Counseling and People Who Stutter contains an excellent description of the therapeutic alliance and other aspects of helping the person who stutters change his behaviors, thoughts, and feelings.
Proceedings of the NINCD workshop on treatment efficacy research in stuttering, September 21–22, 1992. (1993). Special Issue of Journal of Fluency Disorders, 18.

This issue of the journal contains chapters by more than a dozen specialists in stuttering treatment and research. Each chapter deals with an area related to treatment efficacy. Although somewhat out of date, these chapters are good examples of the kind of literature reviews that need to be redone every two or three years.
Shapiro, D. (2011). The clinician: A paragon of change. In Stuttering Intervention: A Collaborative Journey to Fluency Freedom. Austin, TX: ProEd.
This textbook contains two excellent chapters on clinician characteristics. One deals with the “magic” of the client-clinician relationship and touches on many of the attributes needed by effective clinicians. Another discusses the processes of students becoming qualified clinicians. The author touches on many aspects of supervision and analysis of clinical interactions.
Van Riper, C. (1958). Experiments in stuttering therapy. In Jon Eisenson (Ed.), Stuttering: A Symposium. New York: Harper & Row.
This chapter describes the stuttering treatments that Van Riper experimented with in his first 20 years after leaving Iowa and setting up a speech clinic at what became Western Michigan University. Of particular interest is the systematic changes he made in treatment protocols year by year to develop the most effective methods and the five-year follow-ups he made to measure long-term progress.
Van Riper, C. (1975a) The stutterer’s clinician. In Jon Eisenson (Ed.), Stuttering: A Second Symposium. New York: Harper & Row.
This chapter is still a useful description of the attributes that may be important in clinicians who treat stutterers. It also contains excellent sections on clinicians’ roles in motivating clients and discusses the subject of whether clinicians who themselves stutter should treat clients who stutter.
Zebrowski, P. (2007) Treatment factors that influence therapy outcomes of children who stutter. In E. Conture, & R. Curlee (Eds.), Stuttering and Related Disorders of Fluency. New York: Thieme.
This chapter has some excellent summaries of research relating to variables that can affect therapy outcome, including information about the relationship between the client and clinician.
Treatment of Stuttering in Younger Preschool Children: Borderline Stuttering

An Integrated Approach
Author’s Beliefs
Nature of Stuttering
Speech Behaviors Targeted for Therapy
Fluency Goals
Feelings and Attitudes
Maintenance Procedures
Clinical Methods
Clinical Procedures: Indirect Treatment
Severity Ratings
Baseline Speech Measures
Family Interaction Patterns
Slower Speech Rate with Pauses
Teaching Slower Speaking Rate with Pauses
Trying Slower Rate with Pauses in the Clinic
Using the Slower Rate with Pauses at Home
Monitoring Parents’ Practice of Slower Rate with Pauses
Working with Other Aspects of the Parent-Child Interaction
Changes in Family Routine
The Course of Treatment
Supporting Data
Clinical Procedures: Direct Treatment
Direct Treatment for Mild Borderline Stuttering
Direct Treatment for More Severe Borderline Stuttering
Modeling Easy Stutters
The Child Begins Active Participation: “Catch Me”
The Child Begins Active Participation: Play
The Child Produces Intentional Stutters
The Child Changes His Own Real Stutters
Other Clinicians
Edward Conture and Colleagues
Specifics of Treatment
Children’s Group
Parents’ Group
Supporting Data
Palin Centre Parent-Child Interaction
Supporting Data
Stuttering Foundation Approach
After studying this chapter, readers should be able to:
• Understand the difference between indirect and direct treatment
• Be able to plan and carry out indirect and–if needed–direct treatment of ayounger preschool child
• Learn about data collection in the clinic by the clinician and at home by theparents, both of which can be used to guide treatment
• Learn the basics of treatment approaches to preschool children advocated bythree other groups of authors
Younger preschool children: children between 2 and 3.5 years
Indirect treatment: therapy that involves alleviating stresses that the child might be experiencing in communication at home and in other situations
Direct treatment: therapy that works directly on the child’s speech by having him speak more fluently, stutter more easily, or both
Spontaneous fluency: A child’s natural fluency that occurs without work or thought on his part
Maintenance: The process of fading treatment while continuing to support the
child and family so that fluency achieved in treatment does not diminish
Severity ratings (SRs): Numbers on a 1-10 scale given daily by parents to describe their child’s fluency. The SR scale is shown in Figure 8.4 and described in detail in Chapter 8
One-on-one time: A period of about 15 minutes each day during which one parent is alone with the child and follows the child’s lead in play and conversation. In this time, parents can practice new behaviors such as using a slower speech rate with pauses, and children can experience their parents’ full attention
Percent syllables stuttered (%SS): This is one measure of stuttering frequency that is often used as data to determine how much a child is stuttering at a particular time. We suggest using %SS along with SRs as an indication of whether treatment is working
Easy stuttering: A very mild type of stuttering characterized by slow and relaxed repetitions or prolongations of sounds that are brief and relaxed. This can be a target of direct treatment, with the goal being to change stuttering into normal disfluency
Playing with stuttering: Pretending to stutter in a way that deliberately changes some aspect of the stutter, such as how many repetitions are produced. This activity is thought to decease the child’s frustration and fear of stuttering and thus reduce tension and struggle
In previous editions of this book I have described the stuttering in younger preschool children (2 years–3.5 years) to be at the “borderline” level with loose and relaxed repetitions. Treatment for them should be indirect, aimed at changing the environment. However, a few children this young are more advanced in their stuttering; they are starting to add tension to their stuttering and may be aware of and frustrated by it. But I would refrain from direct treatment aimed at having them learn to change their speech. Almost all of these children between 2 and 3.5 years have been stuttering for less than a year, and many have a likelihood of natural recovery. Therefore, everything must be done to avoid interfering with natural recovery and instead to facilitate it. It is also important to realize that most children between 2.5 and 3 years aren’t cognitively ready to learn from direct treatment how to change their stuttering (Fig. 11.1).

Figure11.1 Treatments for younger preschool children who stutter.
Most treatment of stuttering at this age is indirect because it involves working with the family environment to decrease stress and to increase fluency, rather than working directly on the child’s speech. The initial focus is decreasing the family’s concern, trying to understand their feelings, and helping them change selected aspects of the family-child interactions. If a child’s family can discover ways to facilitate the child’s fluency, they become confident in their ability to effect change and are able to assume long-term responsibility for the child’s fluency. If this is not effective, then direct work on the child’s speech by the clinician with some help from the family is appropriate.
In this chapter, I will use the terms “family” and “parent” or “parents” interchangeably to indicate the important adults with whom the child commonly interacts. Cultures and families differ in who should be involved in the child’s treatment. When I imply that one parent is the major player in the interaction patterns I describe, the reader should freely adapt the treatment to suit each situation–one parent or two, older children or cousins, nannies or grandparents.
I will illustrate our approach to treatment with the case example of Ashley, the 2.5year-old preschool child whom we introduced in Chapter 1.
Author’s Beliefs
Nature of Stuttering
As I described in Chapter 7, stuttering at onset usually occurs as a result of the interplay between a child’s constitutional predispositions and the stresses resulting from developmental demands and the environment. Treatment for stuttering in younger preschool children is based on the assumption that when stresses on the child and his speech can be decreased, his stuttering will taper off, and he will become normally fluent. I believe that the plasticity of normal neural maturation allows most of these children to compensate for constitutional predispositions toward stuttering. For such flexibility in development to blossom into normal fluency, however, the clinician and family must provide an environment that fosters fluency and diminishes negative experiences with speaking. And this must be done promptly. If too much time passes, the child may become negatively aware of his stuttering and become frustrated by it. When this happens, frustration combined with concern about negative listener reactions may motivate the child to develop escape and even avoidance reactions. These more advanced forms of stuttering are usually more resistant to treatment.
With the borderline stuttering of younger preschoolers, I seldom treat the child directly, at least not at first. Instead, I work with the child’s family to help them reduce environmental stresses. Stress is normal in the life of every child, but the child with borderline stuttering may simply be more vulnerable to fluency breakdown under normal stresses. As you will see in my description of clinical procedures, I usually begin by informing and educating family members about ways they can reduce stresses and foster fluency. I demonstrate a facilitating style of communicative interaction as a model for the family and meet with them once a week to support and guide their efforts in finding ways to help their child.
If indirect therapy is not effective in reducing stuttering after six weeks, or if the child’s stuttering proves to be more advanced than initially thought, I add more direct procedures. My direct approach for stuttering in younger preschoolers consists of a hierarchy of activities that focus on playing with stuttering and changing it to a milder form.
Speech Behaviors Targeted for Therapy
Because I don’t treat the child’s speech directly, none of the child’s speech behaviors are specifically targeted for direct change. Instead, the family’s interaction styles, including both speech and nonspeech behaviors, are the focus of treatment. For example, I help family members learn to speak in a slow and relaxed manner, and I support their efforts to make other aspects of their interactions with the child who stutters as nonstressful as possible. In those rare instances when this approach doesn’t soon decrease stuttering, the child’s repetitions and prolongations are targeted for change.
Fluency Goals
I believe that children who stutter at the borderline level who have no serious concomitant problems can achieve spontaneous fluency. With effective early intervention, this goal is readily achievable because the child’s maturing nervous system gradually increases his capacity for fluent speech.
Feelings and Attitudes
The main focus of treatment is on the behaviors of family members and others who interact frequently with the child. Consequently, the child’s feelings and attitudes are not dealt with directly. However, as the family and I monitor the child’s fluency, we try to ensure that he is not developing negative attitudes about speaking or about his disfluencies. If the child’s stuttering persists and shows periodic worsening, he may become more frustrated by it when it is at its worst and may soon acquire the escape behaviors (like increased tension in stutters) seen in more advanced stuttering. In these cases, the addition of more direct intervention is used to deal with such feelings.
Maintenance Procedures
Many younger preschool children achieve fluent speech soon after their families have made some environmental modifications, and most maintain fluency without further treatment. However, it is important to keep in contact with the family even after formal treatment has stopped to prevent their reverting to old, more stressful interaction patterns. This maintenance, through telephone calls or e-mail, is gradually faded.
Clinical Methods
Working with stuttering at this level involves a variety of therapy procedures. I educate families by providing them with DVDs and reading material (such as that from the Stuttering Foundation [(SF)]) to help them understand the nature of stuttering and the ways in which they can help the child become more fluent. I counsel families by listening to their concerns and trying to understand their hopes, desires, fears, and frustrations. I brainstorm and problem solve with families when I help them choose aspects of their interaction patterns to modify. I collect data on the child’s speech and on the family’s perceptions of his stuttering and fluency. And finally, I provide support as the child’s stuttering decreases and the family strives to maintain their new styles of interaction.
Clinical Procedures: Indirect Treatment
This section describes the stages of indirect treatment, including continuing assessment of the child’s speech at home and in the clinic, introduction of a slower speech rate with pauses, and introduction of other changes that the clinician and the family choose.
Severity Ratings
As I explained in Chapter 9 on assessment, during the closing interview the family is given a copy of the Severity Rating (SR) Scale (Fig. 8.4), and its use is explained to them. Although the scale was originally designed for use with the Lidcombe Program, it is excellent for use with families of younger children who will be receiving indirect treatment. As you will remember, it is a 10-point scale that the family completes at the end of every day. Ratings range from 1 = no stuttering, 2 = extremely mild stuttering, all the way to 10 = extremely severe stuttering (which may never be seen in a particular child). At each clinic session, after a baseline of the child’s speech is collected in the first 10 minutes, the clinician and parent compare their SRs for the child’s speech during the baseline measure. Agreement is defined as the parent and clinician’s SR for that sample not differing by more than one point. If the difference between the ratings is greater than a single point, the clinician’s rating is assumed to be accurate, and she discusses her rating with the parent to help the parent better understand the rating system and become “calibrated.” The parent uses the scale to make a daily rating of the child’s speech at the end of every day and brings or e-mails the week’s SR chart to the clinician for discussion in each clinic meeting.

Case Example
Ashley’s stuttering, as you will remember from the video clip and Chapter 1, was characterized by multiple part-word and single-syllable whole-word repetitions. Ashley gave little indication that she was aware of her stuttering. Because her stuttering was quite frequent and gradually worsening, Ashley underwent treatment despite her young age. Indirect treatment was carried out via onceweekly home visits by a clinician with experience in stuttering. The clinician began by playing with Ashley on the floor with a variety of toys including dolls and a dollhouse. The clinician used a slow rate of speech with many pauses, as we describe in this chapter. She did not ask Ashley to change her own rate, but only modeled the rate for Ashley and her mother, who observed. Gradually, the mother took a greater and greater part, but the clinician continued to be part of the play so she could observe the mother’s own use of slower speech rate with more pauses. After each session, the clinician and mother talked about how the mother
was doing. The clinician was careful to praise the mother for what she was doing well before suggesting ways she could improve. In addition to slowing her speech rate and pausing more frequently, Ashley’s mother also learned to turn most of her questions into comments. The clinician took data on the child’s stuttering in each session.
This play therapy continued for about eight weeks. Ashley’s stuttering in the sessions disappeared, and her mother reported that she was more and more fluent in situations outside of the sessions. Recently, we video recorded an interview with Ashley at age 8, completely fluent and having no recollection of ever having stuttered
Baseline Speech Measures
At the beginning of each clinic visit, the clinician video records and observes the first 10 to 15 minutes of parent-child play. Attending to both the child’s speech and the parent’s interaction style, the clinician’s first task is to determine an SR for the child’s speech in this parent-child play period. As noted, this will be compared with the parent’s SR for the same sample. The clinician also notes aspects of the parent-child interaction, especially those on which the parent has been asked to work.
Family Interaction Patterns
As described in the section on assessment of the younger preschool child in Chapter 9, my approach to treatment involves study of and experimental change in parent-child interaction patterns. The change is described as experimental because not every change will have positive effects on the child’s fluency; therefore, a change may be tried and discarded if the result of the “experiment” isn’t positive. Here is a list of some conversational interaction patterns observed in some families that may be changed to facilitate the child’s fluency.
1. High rates of speech
2. Rapid-fire conversational pace (lack of pauses or too-short pauses between speakers)
3. Interruptions
4. Frequent open-ended questions or questions that demand an answer
5. Many critical or corrective comments
6. Inadequate or inconsistent listening to what the child says
7. Vocabulary far above the child’s level
8. Advanced levels of syntax
Table 11.1 gives some detailed suggestions for evaluating these parent-child conversational variables. Table 11.2 provides ideas for changes parents can make to facilitate their child’s fluency. Rather than give them this list and ask them to change several things at the same time, I am presenting the list to let the reader know about major areas of communicative interaction to consider with each family.

Table 11.1 Suggestions for Quantifying Family-Child Interaction Patterns

Table11.2 Things Families Can Do to Help the Younger Preschool Child Who Stutters

Although other clinicians may begin treatment focusing on other family-child interaction variables, my own preference in most cases is to begin by helping families reduce their speech rate and increase their pausing. Supporting data on the effectiveness of parents slowing their speech rates and pausing more often is provided at the end of this section. In the next few pages, I will describe how to help parents slow speech rate and increase pause time.
Slower Speech Rate with Pauses
For most families, the clinician starts by helping the parents reduce their speech rate and increase their pause time when talking with their child. The evidence reviewed in Chapters 2 and 3 suggests that individuals who stutter have constitutional deficits that make it challenging for them to produce speech at rapid rates. If parents provide a model of slower speech with plentiful pauses, this model alone will probably influence children to speak more slowly (Guitar & Marchinkowski, 2001). More importantly, a model of slower speech with adequate pauses has been shown to reduce children’s stuttering (e.g., Guitar, Kopff-Schaefer, Donahue-Kilburg, & Bond, 1992; StephansonOpsal & Bernstein Ratner, 1988; Zebrowski, Weiss, Savelkoul, & Hammer, 1996). The clinician emphasizes that parents should not tell the child to slow his speech rate; such direct instruction tends to be ineffective and to annoy the child. Simply speaking in the style of television’s Mr. Rogers will have the desired effect. Parents and clinicians can refresh their memory of this speaking style by watching YouTube videos of Mr. Rogers, including his acceptance speech at the 1997 Emmys.
Teaching Slower Speaking Rate with Pauses
After rehearsing a slower speech rate with pauses, you can meet with the parents, model this style of speaking, and ask them to try it. Sometimes beginning with a reading passage is easier than conversation. Most parents find it slightly embarrassing to speak this way at first, but your modeling this style will make it easier for them. Strongly reinforce them for the things they are doing well. Once they have a pretty good style, if you can video or audio record their speech and play it back to them, their experience of hearing and watching themselves speak this way will help them remember it. If you have a laptop with a camera, you can record a clip of them speaking this way and e-mail it to them so they will have it at home to refresh their memories from time to time.
Trying Slower Rate with Pauses in the Clinic
After the parents feel comfortable using the new speaking style, the clinician and a parent can carry out some play interaction with the child and use the slower rate with pauses (Fig. 11.2). If the child asks the parent why he or she is talking in a funny, slow way, the clinician or parent may explain to the child that the parent talks too fast and needs to learn to slow down. With some children, we enlist them to remind the parent to slow down if they think the parent is talking too quickly. Children delight in correcting adults, especially their parents.

Figure11.2 Clinician models interaction patterns while the mother observes
Using the Slower Rate with Pauses at Home
If the clinician is not satisfied with the parents’ ability to speak with the child using this new speaking style, home practice should be delayed until the parents have mastered it. However, most parents pick up the new style quickly and they can begin using it at home immediately. One parent should try to spend 15 minutes a day playing alone with the child and using the new speaking style. The best time is in the morning because it may influence the child for the rest of the day. Many families are too busy at this time, feeding and dressing their children and themselves; for them, one-on-one slow speech practice in the morning may be possible only on weekends. In this case, any 15 minutes per day of one-on-one time with the child is acceptable. Most important is that the parents do it every day. If one parent does most of the one-on-one play with slow speech, the other parent should also use the slower speech rate when talking with the child whenever he or she can.
Monitoring Parents’ Practice of Slower Rate with Pauses
Most parents benefit from consistent support for any changes they are making in their interactions with their child. I often keep in touch with them between clinic visits via e-mail or telephone, but I always ask them to keep a journal of their experience with the new speaking style by making notes on the SR chart they are completing every evening. When parents are willing and able to video record themselves at home using the new speaking style during their 15-minute daily interactions, it is a fine motivation for them to practice and a good way for me to monitor their progress. In any case, the interaction they have with the child at the beginning of each session allows me to be sure they are using an effective speaking style as well as allows me to collect data on the child’s percent syllables stuttered during this interaction. When I’m observing this, I can also read the SR chart with the parents’ notes about their daily work between clinic visits. In the discussion that follows the parent-child interaction, I am careful to reinforce the parents for everything they are doing well. In all discussions, the clinician’s role is first and foremost to be an empathetic listener, allowing the parents to take the lead in assessing their progress and formulating plans to work on change.
Working with Other Aspects of the Parent-Child Interaction
As the clinician works with the parents or other family members on their speech rate and pausing, she continues to assess the child’s progress toward normal fluency, as indicated by percent syllables stuttered (%SS) in the clinic, SRs at home, and discussions with the family. If these indicators of fluency do not indicate a steady downward progression of stuttering in the first three or four weeks of treatment, the clinician and family should consider other aspects of the parent-child interaction that may be putting pressure on the child’s fluency. Consider items 3 to 8 in Table11.1.
Remember that most of the interaction patterns in families of children who stutter are not abnormal or particularly negative. They usually are quite typical of the culture in which the child is being raised. However, a child sensitive to communicative pressure may benefit from some modification of family communication patterns in ways that facilitate his fluency. It is vital that the clinician help the family understand that they are not causing the child to stutter because of inappropriate communication patterns. Instead their communication is normal, but they can help their child by changing a few aspects of their communication to facilitate fluency.
One of the first things I do if I sense some aspect of the family communicative style might need a little modification, is to have the parents watch the video recording of their interaction with their child with me. As we watch together, I praise many things that the parent is doing well. Praise is vital to develop the parents’ confidence and to encourage them to keep doing things they are doing well. As I praise the parents, I listen acceptingly to their comments even if they are self-critical. We watch together for things that a parent and I both feel might be putting pressure on the child. The best situation is when a parent notices something to change–something that I also feel may be pressuring the child’s fluency. We then together plan to change that aspect of the interaction and observe the results.
For example, in the third week of treatment, a parent was doing a great job using a slower speech rate with pauses. However, the child’s fluency–which had increased somewhat–had now plateaued. The parent and I watched the most recent video recording, and as we watched, I praised her slow speech with pauses. After a few minutes, the parent commented that she was surprised to see that she asked her child so many questions, rat-tat-tat, one right after the other. I agreed with her and we discussed alternatives, such as making comments instead of asking questions, and she tried this out during the week. The following week, the child’s SRs showed further increases in fluency, and the parent’s interaction at the beginning of the session revealed an impressive decrease in questions. This change, accompanied by the slower speech rate with pausing, was enough to increase the child’s fluency to normal levels, which was maintained long term.
Changes in Family Routine
In addition to changing conversational interaction patterns, a family may identify other stresses on the child that need to be changed, such as the amount of individual attention the child receives and the “busyness” of the family’s schedule. My main function in helping families work on such stresses is to give them information about areas of changes that others have found helpful and to be a sounding board for their plans for changing. I encourage them to assess, informally, the effects of these changes on the child’s fluency and his overall adjustment. Although my praise and appreciation may help, a significant change in the child’s stuttering is the real motivator. Notes the parent makes on the chart of daily SRs will help you identify factors that may facilitate fluency or cause upward spikes in a child’s stuttering. A parent, for example, noted on her SR chart that her child’s stuttering flared up if she left the room while he was playing. She alleviated this stress by being careful to let the child know ahead of time if she were about to leave the room and that she would be right back.
This example is a reminder of the importance of parents’ attention for a child’s selfesteem. When a child senses that his mother or father understands him and genuinely cares about him (cares about what he likes to do, what he thinks about things, and how he feels), the child feels more comfortable with himself, is less anxious, and is better able to speak easily. For many younger preschool children who stutter, a little more one-on-one time spent with a parent each day, preferably in the morning, can boost fluency tremendously. Although the morning can be the most difficult time for parents who work outside the home, one-on-one fluency time in the morning can have a positive effect on the child’s speech for the rest of the day. If mornings are too difficult, some one-on-one time in the afternoon can also be very helpful. The time does not need to be long, just 15 to 20 minutes, but the parent needs to be with the child in a place where they won’t be interrupted.
The child should choose what to play or talk about, and the parent should follow the child’s lead, participating as the child directs. As a parent becomes more and more comfortable with this nondirective play, he or she may want to explore ways of helping the child feel really understood. One of the parents we worked with, for example, learned to “mirror” her child’s momentary emotions as they built a tower of blocks together. When the child placed a block on the tower and it fell off, she would quietly murmur a sound of disappointment, echoing the child’s facial expression. This child made impressive gains in fluency in only a few weeks, and I believe that this parent’s deep attention to the child may have contributed significantly to this change. Although not every parent could be expected to achieve the level of empathetic response that this mother did, increased caring attention is probably a realistic goal for most families.
Attentive play can become child-directed conversations as a child grows older, and such conversations can continue the process of helping the child develop a sense of being loved, understood, and appreciated. In his article Making Time for Your Child, the child psychiatrist Stanley Greenspan suggests “In spontaneous, unstructured talk or play, try to follow your child’s lead. The goal is to ‘march to your child’s drummer’ and to tune in to the child at his level” (Greenspan, 1993, p. 111).
The Course of Treatment
Sometimes families report that their attempts to make changes have been fairly successful. For example, they may have been able to slow their speaking rates and to simplify their language and may have seen improvement in their child. I let them know that their changes have been key factors in the child’s improvement and stress the importance of continuing them. It is easy to resume old patterns after some improvement occurs, whether it’s the challenge of losing weight or helping a child become more fluent.
Each child and each family is unique in how they respond to treatment, but it is possible to note some common trends. For example, some children become much more fluent soon after the family makes one or two changes in their environment. Occasionally, a child may become fluent immediately after an initial session, possibly because the family is less anxious about his disfluencies after sharing their concerns with a professional. Whatever the cause, early and immediate fluency gains should be viewed with cautious optimism. I share the family’s pleasure at such dramatic change but suggest that their child’s fluency may be fragile and will need to be nurtured by our continued efforts to create a facilitative environment.
Sometimes the path toward fluency is rough and irregular. The child may make little or no progress or may improve for a while and then return to his old pattern of disfluency. When this happens and the family or clinician feels frustrated by slow progress, further exploration of the family’s feelings about the child’s stuttering is called for. Many times family members worry about the child’s future, afraid that stuttering will be a serious handicap for him. Sometimes there is lingering guilt about having caused his stuttering. Often it is hard for parents to accept the blemish they feel that stuttering creates on the family image.
Whatever the source of a family’s anxieties, their concern about stuttering may easily radiate to the child in their reactions to his stutters. Unwittingly, family members may show their anxiety or disappointment through facial expressions or body language, which may make the child “hesitate to hesitate” and thus stutter more severely. Open and frank discussions with the family about their feelings and concerns are likely to be more helpful at this point than trying to change their reactions. In such discussions, the clinician’s role is to make it easier for the family to talk about their concerns, so I listen carefully, try my best to understand them, and convey my understanding with acceptance and respect. When family members feel understood and accepted, it is easier for them to share their feelings and accept them. When this occurs, some feelings may change, and in turn, the child’s stuttering may decrease, possibly because his stuttering no longer seems so terrible to the family.
Another barrier to changing a family’s interaction patterns is the fact that some styles of interaction reflect important cultural values. For example, in the urban eastern United States, family members sometimes finish each other’s sentences, conveying a closeness and solidarity within the family that is highly valued. If they are asked to speak more slowly and pause between speakers’ turn takings, such changes would conflict with one of the family’s implicit cultural values. Another example might be parents who frequently teach, correct, and criticize their children’s behavior. This “instructional” mode of interaction may reflect the importance that the family’s culture places on education.
I believe it is important to explore how the family feels about changes they are considering. In some cases, they can find ways to change other variables that will be as effective, thereby leaving unchanged those interactions that are of value to the family. Several years ago, I worked with a parent who spoke very rapidly to her child who was showing some borderline stuttering. She resisted changing her speech rate because “it isn’t the way we talk.” In addition, she was frequently critical of her child’s behavior. Consequently, I encouraged her to use positive reinforcement for fluency, as described in the treatment of beginning stuttering in Chapter 12, which I adapted from Onslow, Andrews, and Lincoln (1994), and asked her to let her child know with upbeat statements of praise that she liked his smooth fluency. The child’s stuttering diminished almost immediately, and she was delighted with her ability to help her child.
Sometimes a family may resist change and doesn’t fully participate in treatment. There may be psychological issues that need to be resolved through referral to a family counselor, or the family may have other, more serious problems with which to cope. In such cases, I talk with the family directly about my concerns. This usually leads to an open discussion of their situation, a referral to a family counselor, or in rare cases, their decision to withdraw the child from stuttering therapy for the time being. If this happens, I let the family know that I remain available to them, and I try to stay in contact by occasional phone calls or e-mails to make it easier for them to resume the child’s therapy if they wish to.
Indirect treatment of a younger preschool child is often effective within five or six sessions, over a period of one or two months. The child’s speech becomes markedly less disfluent. Part-word repetitions become whole-word or phrase repetitions, which are more like typical children’s disfluencies, and the family’s concerns about the child’s speech diminish. When this happens, I review the changes the family has made with them and the changes in their child’s stuttering that reflect his improvement. Using this information, I help the family develop a plan to deal with periods of increased stress that may prompt stuttering to reappear. Most families feel that they have a handle on how to reduce stress on their child at this stage of therapy, and their experiences in observing and changing their behaviors have given them confidence. If their child’s stuttering suddenly increases, they know how to examine their speech rates or attentiveness when talking to the child and how to examine other aspects of their interactions and implement needed changes.
Effective maintenance for stuttering in younger preschool children is the result of two things: (1) helping the family to view the child’s stuttering more objectively with less anxiety, guilt, or panic, and (2) building the family’s confidence in their own ability to implement problem-solving skills they’ve learned to use when the child’s disfluencies increase. Sometimes, however, despite a family’s best efforts to respond constructively, stuttering returns. This may occur after an increase in stress from some trauma or from normal life events, such as moving to a new house, or it may accompany a growth spurt in the child’s language. On the other hand, it may be unexplainable. Whatever the cause, the family should feel comfortable getting back in contact with the clinician. I let each family know at the end of therapy that relapse is possible, not abnormal, and that I would look forward to seeing the child again if help is needed.
Supporting Data
Many years ago, my colleagues and I published a study that evaluated the effect of changing parent-child interactions with a 5-year-old child who stuttered (Guitar, 1978; Guitar et al., 1992). Although this child was an older preschooler, the principles of working with the family on their interaction style were similar to those described for the younger preschool child. Our approach to treatment was to video record parentchild interactions over five treatment sessions and then view the videos with each parent. When viewing the videos, we let the parent decide what to work on in the intervening week and then recorded a new parent-child interaction after a week of work on changing the behavior they had selected. After six sessions, the child’s stuttering had diminished to the level of normal disfluency; we followed the child for 10 years, and the stuttering never reappeared. In an analysis of the parents’ behavior and the child’s stuttering, we discovered that the changes in parent behavior most related to the child’s improvement in stuttering were the mother’s reduction in her speech rate and her becoming more accepting in her comments.
A variety of other studies have shown that changes in parent’s communicative interactions affect their children’s stuttering. Stephenson-Opsal and Bernstein Ratner (1988) demonstrated that when the mothers of stuttering children slowed their speech rates, the children’s stuttering decreased. Starkweather, Gottwald, and Halfond (1990) reported on 29 children they treated for an average of 12 sessions (some required as many as 40 sessions), all of whom completely recovered. Their approach involved primarily modification of the parents’ behavior, including reduction of speech rate, having special speech time, matching parent language to child language, and reducing parents’ negative reactions to stuttering. Zebrowski and colleagues (1996) showed that decreases in mother’s speaking rate and pause time were associated with decreases in stuttering in some children.
Further supporting data on this approach are presented in the outcome measures of the Michael Palin Centre’s treatment of preschool children later in this chapter. Moreover, a report by Franken, Kielstra-van der Schalk, and Boelens (2005) provides evidence of the effectiveness of parent-child interaction therapy. In the next chapter on treatment of older preschoolers, the treatment approach described by Gottwald (2010) uses a great deal of indirect treatment, but supplements it with direct treatment when needed. She reports that 26 of 27 children were speaking normally a year or more after treatment ended.
Clinical Procedures: Direct Treatment
My approaches to therapy evolve as I learn more about children who stutter and treatment options. Recently my choice of a direct treatment approach for those borderline stutterers who have not responded to indirect treatment has been the Lidcombe Program, which I describe in detail in Chapter 12. The material in this section on direct treatment has been helpful for younger preschool children who need something more than indirect treatment, and I would recommend it to those clinicians who do not choose to use the Lidcombe Program or who have not yet been trained in it.
I don’t use direct treatment with every child who is a borderline stutterer, but it is a powerful alternative when indirect treatment does not decrease the child’s stuttering after six weeks. The causes of failure with an indirect approach are often unknown. Sometimes, a family seems unable to modify the child’s environment as planned, or they do, but the child’s stuttering persists unchanged or increases. In these few cases, if the child’s disfluency remains at the borderline (rather than beginning) level, I try a slightly more direct approach, as described in the next section.
Direct Treatment for Mild Borderline Stuttering
Most younger preschool children with borderline stuttering are only slightly aware of their disfluencies. Their repetitions appear relaxed, and they show no signs of extra effort or attempts to “fight” their stutters. They also are normally fluent a great deal of the time, and I think they have the capacity to develop entirely normal fluency. Consequently, when I use direct treatment for mild borderline stuttering, I focus on the child’s fluency, assuming that he will easily be able to increase the amount of fluency he has and “outgrow” his stuttering with our help. I follow much of the behavioral management strategies used by the Lidcombe Program, which is described in Chapter 12. I train parents to respond to fluency with praise, and unlike the Lidcombe Program, I ask them to ignore stuttering unless the child is momentarily distressed by a stutter, in which case, I suggest the parents comment acceptingly on it. Thus, it is not, strictly speaking, a Lidcombe approach.
I usually begin by training one of the child’s parents to use praise for fluency during the daily one-on-one time with the child. The parent might say, “Gee, that was really smooth talking” or “I like the way you said that.” The clinician and parent should decide how frequently to use positive reinforcement, but most children are annoyed by praise if the parent gives it too often. A good ratio to begin with is one praise for about every fifth fluent utterance. These fluent utterances do not need to be consecutive. A few children are annoyed by any praise at all given by the parent. In this case, the parent and the clinician can talk with the child about using something besides typical praise. Some children prefer their own phrase. One child wanted his parent to say “That was good monkey talk!” Another child asked for a gesture (thumbs up) instead of words.
As in parent-child interaction therapy, parents keep daily logs of the child’s overall fluency for each day, using the 1-to-10 SRs described earlier. When the child has made substantial progress in decreasing severity, the clinician guides the parent in gradually replacing praise for fluency in the daily one-on-one sessions with praise used occasionally during other activities during the day. While the parent is carrying out this direct therapy, it is critical for him or her to attend weekly meetings with the clinician to demonstrate using the procedure, to share SRs, and to discuss progress and problems. It is also important for the family to continue one-on-one sessions with the child and to continue the changes made in their interactions and family lifestyle.
Direct Treatment for More Severe Borderline Stuttering
Some children with borderline stuttering are beginning to have negative feelings about their disfluencies but are not showing the full-blown signs of physical tension or escape behaviors that characterize beginning stuttering. Still, they may occasionally express real frustration with their stuttering.
Typically, I work with children having more severe borderline stuttering for about 45 minutes each week. I also continue to provide encouragement and support to the family in helping them make the child’s environment as facilitating to fluency as possible. Our direct treatment activities are presented in a hierarchy that the clinician and child ascend as far as is necessary to bring the child’s disfluencies into the range of normal. Progressive steps are taken when the clinician senses that a child is feeling competent at the current step. Thus, progress may be rapid or slow or sudden or gradual, depending on the child’s feeling of comfort and mastery with the tasks at hand. There is no need to hurry this process. It should take place within the context of games and activities that make the focus on stuttering casual. The clinician needs to remain alert to the child’s immediate sense of confidence and self-esteem in selecting the moment to move the child to the next step in the treatment hierarchy.
Modeling Easy Stutters
I begin direct treatment rather indirectly by providing models of easy stuttering in my speech. If the child’s repetitions are fast and abrupt, my models are slow with gradual endings. If the child has many repetitions or long prolongations, I repeat or prolong sounds briefly. These models are done casually during play with the child. I don’t produce them immediately after the child stutters but insert them randomly, about once every two or three sentences, as if I were stuttering as I talked.
Once the child has become acclimated to the models of easy stuttering after 10 or 15 minutes of play, I begin to make accepting comments about them. I might say, for example, “Hmmmm, I used slidey speech on that word, didn’t I?” or “That word stuck a little, but that’s OK, I slid right out of it.” Most children appear to be shyly interested in what I am talking about, and direct therapy can continue to develop. A few, however, may react negatively and say such things as, “Don’t do that!” or “I don’t like it when you do that.” For them, direct therapy needs to proceed slowly to allow my acceptance and support during play activities to gradually counteract the child’s anxiety.
If the child has begun to experience the first pangs of frustration from stuttering, which can be inferred from his questions or complaints about getting stuck on words, I will try to help the child express this. Even though I am making comments that show acceptance of my own stuttering, I occasionally may produce a longer than usual stutter and say, “Sometimes they go on for a long time. That feels weird.” I continue to try to sense what the child is feeling and to empathize as naturally as possible. I use this empathic focus not only when I am modeling easy stutters, but throughout direct treatment.
For children who evidence periods of acute frustration with their stuttering, parents should be coached on how to make empathetic statements in a calm, soothing, slow style when the child is going through a difficult time. As I do direct therapy, I try to involve the parents in appropriate activities both at home and in the clinic. If their indirect treatment has not been effective, I need to be sure they do not feel pushed aside by my direct therapy.
The Child Begins Active Participation: “Catch Me”
When I sense that a child is comfortable with my easy stuttering models, I see if the child will take part. I may say, for example, “Can you help me? Sometimes when I get stuck on a word, it goes on and on. Then I try to make my stuck words real slow and loose, and it helps me get unstuck. But sometimes I forget. If you hear me go on and on like thi-thi-thi-thi-thi-this, just say, ‘There’s one,’ and I’ll try to make it slow and loose with slidey speech.” When the child catches me, I will change a fast, tight repetition to a slow, loose one. As I model stuck words, I choose a style of stuttering similar to the child’s.
Praise should flow liberally when the child catches one of my modeled stutters. This provides the child with an initial sense of competence that is associated with something he previously felt to be out of control. For many children, tangible rewards, such as small snacks or turns at a game, are important motivators and should be used along with praise to establish the child’s ability to catch the clinician’s stutters.
The Child Begins Active Participation: Play
This stage can either follow or precede “Catch Me.” It depends on the clinician’s judgment about which activity would be more comfortable for the child. Sometimes you may start one of these stages but find the child is not ready and you switch to the other. The playing with stuttering stage engages a child in following the clinician’s lead in playfully imitating disfluencies that are similar to his own, such as repeated or prolonged sounds. The purpose is to desensitize the child to the frustration that sometimes arises in borderline stuttering. It is a process that may take place because play can give a child a sense of mastery without the risk of failure. The concept of play is quite interesting. Scientists speculate that children’s play is an opportunity for them to practice and master skills that are needed in adulthood. Playing with stuttering may take advantage of children’s natural tendency to play and provide them with the pleasure of mastery and control over something that has been frustrating and sometimes even frightening.
Take, for example, the child who stutters primarily in a repetitive fashion. The clinician might say, “Let’s play a game of saying some sounds over and over and see how many times we can say them. I bet I can say a sound five times! Watch this. Baba-ba-ba-ba! Can you do it five times?” Or it can begin by making sounds for animals, puppets, or other toys: “Hey, this is a zebragella! It goes ‘llllllla! lllllla!’ (using prolongations). Then it jumps around like this (clinician jumps around) and chews carpet (clinician pretends to chew the carpet).”
The clinician and child can keep incorporating such play into their routine as long as the child finds it fun and the clinician can free herself to enjoy uninhibited play. From playing with repeated or prolonged sounds, the clinician can build a bridge to playing with repeated or prolonged sounds in conversation and in time, to the child’s actual stutters.
The Child Produces Intentional Stutters
After the child is able to catch the clinician’s stutters and appears comfortable doing it, the clinician should begin looking for opportunities to ask the child to produce a stutter intentionally. This can be done most easily by pretending to have trouble producing slow, loose stutters. For instance, the clinician might say, “I can’t seem to make this one slow and loose. Can you show me how to do it?” Again, this should be done intermittently and casually mixed in with other activities that are fun for the child.
Praise and, if needed, tangible rewards are used to help the child build confidence. When the child is able to produce slow and loose stutters, the clinician can let the parents know, in the child’s presence, about this accomplishment, focusing on the child’s ability to teach the clinician. If the child seems proud of this accomplishment, the clinician can take advantage of this opportunity and have the child show intentional stutters to the parents. This not only desensitizes the child to stuttering with the parent, but it also desensitizes parents to the child’s stuttering and models acceptance of the child’s stuttering for them.
The Child Changes His Own Real Stutters
For many young children whose stuttering fluctuates between mild and severe levels, these direct therapy activities, combined with a facilitating environment provided by parents, may be enough to advance their fluency into the normal range within a few months. For those whose stuttering persists, still another stage of direct therapy may be necessary. In such cases, I look for opportunities when the child seems ready to modify his own stutters.
I begin by responding to a few of the child’s real stutters with accepting comments to help the child feel comfortable with his stutters. I might say, with an accepting voice, “Oh, that one was a little bumpy on ‘my-my-my car…,’ ” and then return to the business of playing. After further play, when the child stutters again, the clinician can model an easier and slower style of stuttering on the same word and comment positively about it. I then ask the child to imitate my easier stutter and praise him for doing so, using reinforcements and guidance to shape his stuttering to a slow, relaxed style.
I look for slightly slower and easier stutters in the child’s speech and reward them. Even if the child intentionally stutters, but in an easier way than he stuttered previously, I reward him. From this point on, the clinician uses a combination of modeling and reinforcement to shape the child’s stuttering. It is the deliberate slowness and “easiness” with which the child produces repetitions or prolongations, along with the sense of playing with stuttering, that make it possible for the child to begin feeling a sense of control. This in turn should reduce his frustration and fear, further diminish tension, and enable him to move through stutters with minimal effort.
After the child is able to make his stutters slower and easier in the clinic, generalization may occur away from the clinic without the need for formal transfer activities. Such “spontaneous” generalization may be a result of the child’s increased self-esteem from gaining mastery over behavior he previously felt uncomfortable about and felt was out of his control. Consequently, emphasis should be placed on the stutters that a child handles successfully, rather than when he loses control.
If generalization is not occurring automatically, I work with family members to make the child’s ability to play with and modify stutters a point of pride at home. Initially, the child can teach parents and siblings to stutter in the clinic under the clinician’s guidance. Then the clinician can work with the child at home and involve family members when appropriate, so that the parents learn to use positive reinforcement selectively to increase the child’s slow and easy stutters and let him know that he is appreciated. Even though the emphasis here is on slow and easy stutters, the effects of speech and language maturation and the increasing confidence that the child feels in his speech as a result of reduced frustration should result in normal fluency.
The approaches of several other clinicians are described here. Many of these approaches are used not only for borderline stuttering in younger preschool children, but for beginning stuttering in older preschool children as well. I have selected them because they all involve the child’s family, which I consider of major importance when working with preschool children. The nature of intervention ranges from monitoring the child’s stuttering to helping parents change their interaction patterns to direct work on the child’s way of speaking, if needed.
Even though some of these approaches present data on their effectiveness, clinicians using any approach should collect their own data on progress and outcome. As suggested in Chapter 8, baseline measures of the child’s stuttering at the beginning of treatment should be made in a valid and reliable way. Because preschool children are highly variable, recordings of the child’s speech should be made at home as well as in a clinical setting. The Stuttering Severity Instrument-4 (Riley, 2009) should be used to assess frequency and severity. When treatment begins, weekly measures of progress should be made; percentage of syllables stuttered in the clinic and daily SRs of speech at home made by a parent are effective and efficient. When the child has achieved fluent speech (SRs at home of 1 [normal fluency] and less than 1 %SS in the clinic), a maintenance program should be started, involving continued measurement at home and during gradually faded clinic visits. Children with borderline stuttering can be expected to achieve stable, normally fluent speech within six months. Clinicians should assess how long a child is in treatment before fluency is achieved and how well the child maintained that fluency a year after treatment ends.
Edward Conture and Colleagues
Conture’s indirect therapy for preschool-age children (Conture, 2001; Conture & Melnick, 1999; Richels & Conture, 2007) is carried out in parent and child groups that meet separately each session and then as a combined group. The parent group observes portions of the children’s therapy and is provided information, suggestions, and opportunities to help them facilitate their child’s fluency outside the clinic setting. The parents and clinician also discuss child-rearing issues that directly affect the child’s ability to receive maximal benefit from treatment. The children’s group helps the child learn the skills of effective communication.
During the initial phase of treatment, the groups meet once per week to establish the child’s fluency. When the child’s stuttering in the clinic is below 5 percent for four consecutive weeks (which usually takes about 12 weeks of treatment, but for some longer), the family then meets less frequently–but still regularly–with the clinician so that the parents can continue new behaviors more independently, and the child’s increased fluency can be transferred to home and beyond. This is followed by a maintenance phase, to be described.
Conture and his colleagues stress the importance of data to guide treatment. Richels and Conture (2010) describe their evaluation procedures and the important relationships between pretreatment information and the outcomes of treatment, for example, understanding that a slow-to-warm-up child may require a longer period of treatment before becoming successful. They typically collect measures of the child’s stuttering frequency at the beginning of every clinic visit and are beginning to use a rating scale on a monthly basis to examine parents’ perceptions of their children’s disfluencies and children’s responses to stuttering beyond the clinic to determine how well treatment effects are generalizing and to guide dismissal.
Specifics of Treatment
Children’s Group
The children’s group begins with conversation led by the clinician. Samples of 100 words of conversation are obtained from each child to provide disfluency data for each session. Activities in the children’s group then begin with rules that foster good communication, which include listening when someone else is talking, taking turns in conversations, and not interrupting. These rules are described to the child verbally and augmented by brightly colored pictures depicting each of the three rules. Children are reinforced for using the rules appropriately. After rules are reviewed, the clinician engages the children in “story time,” during which reading of an age-appropriate story is intermingled with questions of the children that match each child’s language and fluency abilities (e.g., “forced-choice” questions are asked of a child who is fluent only at this level of linguistic demand, while open-ended questions are asked of a child who is ready for that level of demand). Story time is followed by a craft or game activity that blends turn taking and verbal responses consistent with each child’s abilities. The clinician models the interaction strategies with which the parents are being familiarized in their group (see below).
Parents’ Group
Many of the parents’ group activities are focused on improved communicative interactions that dovetail with the turn-taking rules and adaptation of linguistic demand based on a child’s demonstrated fluency, so that parent-child conversations at home increasingly facilitate the child’s development of fluency, related speech and language behaviors, and ease of communication. The parents learn the following strategies over the course of treatment:
• Speaking more slowly (but still normally) to the child
• Adjusting the length and complexity of utterances to meet requirements of the communicative situation–generally talking to the child using shorter, simpler sentences when appropriate.
• Pausing for a second after the child speaks to give the child plenty of time to finishspeaking and to slow the overall pace of conversation.
• Decreasing the number of interruptions and questions when conversing with thechild
• Decreasing the number of corrections of the child’s speech, language, and relatedcommunicative behaviors
Parents in the group are asked first to observe their speaking behaviors with their child, as well as the child’s speaking behaviors, and then encouraged to discuss their observations with the group. Subsequently, parents learn how to make changes in these behaviors and practice them in a single setting, once a day for about five to 15 minutes (depending on the situation and the parents’ comfort level with new strategies). This helps them experience and learn these changes in a relatively controlled setting, such as nightly bedtime rituals, before trying them in more spontaneous situations, such as during conversations at the dinner table. The first new behavior parents are asked to practice is to speak more slowly but normally to the child and to pause for one second after the child finishes speaking before they start speaking. This change is thought to create a speaking environment that facilitates the child’s fluency and overall conversational turn taking. This approach also attempts to produce a communicative environment in which the child is more likely to feel that he doesn’t have to hurry to speak and helps to reduce how often parents are interrupting their child. The parents’ pause after their child talks will also reduce the extent to which parental communicative behavior may encroach on the time during which the child may be planning and producing spoken language. Another important change for parents is to learn to use shorter and simpler sentences, not continually, but when appropriate. Parents’ modeling of this behavior encourages children to decrease the speed of initiating and maintaining speech as well as more easily adjust the length and complexity of their utterances to communicative requirements, enhancing their fluency. Similarly, parents learn to adjust the level of demand of their inquiries of their child according to their observations of the child’s fluency in particular situations (e.g., keeping it simple, forced-choice questions when children are more disfluent, and using more open-ended inquiries when children are more fluent).
Parents discuss and practice other new behaviors on the list and are encouraged to use the group to support each other in planning situations in which they can try out these changes and then share the results of their efforts. Parents are most successful when they don’t try to change all of their behaviors in all situations but work on only one or two at a time in such specific situations as talking with their child while playing a simple board game or playing alongside the child (e.g., building with blocks, playing with Play Doh, dressing a doll, etc.) or reading a bedtime story. It is stressed to the parents that smaller periods of practice, almost every day, are more effective than longer periods of practice once or twice a week.
Throughout treatment, parents are provided with graphic descriptions of how their child is doing (Figs. 11.3 to 11.5). These “therapy graphs” depict three pieces of information: (1) total disfluencies per 100 words, (2) total stuttered disfluencies per 100 words, and (3) total nonstuttered disfluencies per 100 words. These graphs are used during therapy sessions as well as during parent-clinician counseling sessions to help parents understand expected variations in disfluencies and their child’s progress over time, as well as to help the clinician plan short-term and long-term goals for each child.

Figure11.3 Example of a child whose treatment was successful (Richels & Conture, 2007).

Figure 11.4 Example of a child with highly variable performance during treatment (Richels & Conture, 2007).

Figure 11.5 Example of a child not responding to group treatment (Richels & Conture, 2007).
Conture stresses the importance of the fact that once a child is ready for dismissal, treatment is not abruptly terminated but gradually faded. First, the child’s treatment changes from once a week to once every other week, then to once a month, once every three months, and finally, every six months for a year. If stuttering reappears at any time during this period, the child can be brought back into treatment until he regains fluency. Although this schedule of fading is the best approach, the time of a child’s dismissal is sometimes negotiated with the parents. Some parents want to continue with regular treatment longer than may be necessary, which is only allowed if it is believed to be in the child’s best long-term interest (typically as a means of helping the parent(s) to prepare for independence). Other parents want to discontinue treatment after the child has become fluent but before gradual fading of treatment has been completed. In these cases, parents’ wishes are granted, and the door is left open for them to return if necessary. Based on his experience, Conture rejects a one-size-fits-all approach to determining the length of treatment and criterion for termination. Instead, he considers the individual needs of each child, the nature of his problem, the child’s learning history, the parents’ concerns, perceptions of their child’s progress, and the extent of their involvement in therapy, in determining each child’s pace in moving from skill acquisition to maintenance to dismissal (i.e., establishment of normal disfluency).
Supporting Data
Richels and Conture (2007) reported on 32 children (ages 2–9 to 6–0) who stuttered on an average of 10 words per 100 (stuttering-like disfluencies), averaged 21.3 on the SSI-3 (moderate), and whose time since onset of stuttering averaged 15.5 months. All of the children had received at least 12 treatment sessions–enough time to have shown some response to therapy, according to their experience. Results indicated that on average, these children decreased their stuttering-like disfluencies by 31 percent in 12 sessions.
Palin Centre Parent-Child Interaction
The team at the Michael Palin Centre for Stammering Children have developed a treatment for preschoolers who stutter based on the therapy of Lena Rustin (1991). This approach rests on the premise that the vulnerabilities that underlie the breakdown in children’s fluency also make it harder for them to cope with typical adult-child interactions (Miles & Bernstein Ratner, 2001). Therapy uses video feedback to help parents identify interaction styles that support their children’s fluency and then develop these styles in structured practice sessions at home. While the research indicates that changes in interaction style can be associated with increased fluency (Millard, Nicholas, & Cook, 2008; Guitar, 1978; Guitar et al., 1992; KasprisinBurrelli, Egolf, & Shames, 1972; Stephenson-Opsal & Bernstein Ratner, 1988), parent-child interaction styles are not considered to be causing the breakdown in fluency. Indeed some studies have found that interaction styles alter after the child has started to stutter (Meyers & Freeman, 1985a, 1985b; Kloth, Janssen, Kraaimaat, & Brutten, 1998).
A detailed description of the Palin Parent