Listener perceptions of stuttering, prolonged speech, and verbal avoidance behaviors

Abstract

This study examined listener perceptions of different ways of speaking often produced by people who stutter. Each of 115 independent listeners made quantitative and qualitative judgments upon watching one of four randomly assigned speech samples. Each of the four video clips showed the same everyday conversation between three young men, but differed in how the target person spoke. The four ways of speaking comprised: (1) stuttered speech, i.e., a speech containing repetitions, prolongations, and blocks, (2) hesitant speech, i.e., a speech containing verbal avoidance behaviors like interjections and revisions, (3) a mix of both, and (4) prolonged speech learned in fluency-shaping therapy. Quantitative data revealed that listeners perceived a speaker producing hesitant speech as less pleasant, self-confident, and communicatively competent than a speaker producing stuttered speech or prolonged speech. There were no differences between stuttered speech and prolonged speech. Ratings were partly dependent on the listeners’ implicit theory of speaking difficulties, that is, whether they assumed a chronic speech defect or a temporary problem. Implications of these findings are discussed.

Learning outcomes: The reader will: (1) be able to summarize how different ways of speaking produced by people who stutter are connected with different listener perceptions; (2) be able to explain how the listener's implicit theory of speaking problems influences these perceptions; (3) learn about the clinical implications of the data from this study.

Introduction

Recent studies suggest that people who stutter are more socially anxious than other people (Blood et al., 2001, Ezrati-Vinacour and Levin, 2004, Kraaimaat et al., 2002). They report more anxiety in situations where social evaluation might occur than people who do not stutter, whereas there are no differences in anxiety related to physical danger and daily routines (Messenger, Onslow, Packman, & Menzies, 2004). They tend to hide their stuttering (Petrunik and Shearing, 1983, Vanryckeghem et al., 2004) and feel stigmatized (Corcoran and Stewart, 1998, Klein and Hood, 2004). Putting together these findings, people who stutter seem to be afraid of listener's negative evaluations because of stuttering, and to avoid stuttered speech in order to be perceived in a more positive way (Plexico, Manning, & Levitt, 2009). While there is no clear evidence that the processes of anxiety and avoidance cause stuttering directly, their clinical importance is beyond all question (Manning, 2001, Plexico et al., 2005). To better understand these processes, it may be helpful to examine whether the anticipated negative listener evaluations correspond to reality.

This question of how people who stutter are seen by other people has been studied extensively. For some time, research focused on people's opinions about “stutterers” in general (stereotypes). As a group, people who stutter have been shown to be stereotyped as, for example, shy, nervous, and reticent (MacKinnon, Hall, & MacIntyre, 2007). Recently, the question of how a person who stutters is perceived in a specific situation (listener perceptions) has become more important. However, the patterns of listener perceptions have proven to be much more complex and inconsistent across studies than those of stereotypes. Existing empirical work in listener perceptions will be discussed in the following sections, focusing on (1) perceptions of different modes of speaking often used by people who stutter (e.g., stuttered speech or prolonged speech) and (2) different dimensions of perception (i.e., speakers’ personality attributes like intelligence or self-confidence).

The observable speech behaviors occurring during moments of stuttering are often subdivided into core behaviors and associated behaviors. Core behaviors of stuttering are repetitions, prolongations, and blocks. Associated behaviors are, for example, associated body movements, interjected speech fragments, and negative emotions (Bloodstein & Bernstein Ratner, 2008). However, many people who stutter prevent both core and associated behaviors from happening by using a way of speaking learned in therapy, for example prolonged speech learned in fluency-shaping programs.

In most studies of listener perceptions, core behaviors are used as an example of the speech of a person who stutters, not associated behaviors or post-treatment speech. Susca and Healey, 2001, Susca and Healey, 2002 created audiotapes containing repetitions and prolongations. The samples were obtained from a 27-year-old male who stuttered reading a short passage and varied in severity (0%, 5%, 10%, and 15% stuttered words). Panico, Healey, Brouwer, and Susca (2005) used audio and video samples, again containing repetitions and prolongations varying in severity. A person who formerly stuttered simulated rhythmical two-unit-repetitions and 2 s sound prolongations while reading a short passage. Evans, Healey, Kawai, and Rowland (2008) recorded videotapes of a joke told by a 16-year-old male who stuttered. He simulated two-unit-repetitions and prolongations ranging from 0.5 to 2.5 s in predetermined places in the joke, resulting in samples of different severities. A common finding of these studies was that the more dysfluencies people who stuttered produced, the less they were perceived as competent speakers (for a more detailed discussion of the findings, see Section 1.2).

While these four studies compared perceptions of core behaviors with perceptions of fluency, Manning, Burlison, and Thaxton (1999) were interested in the perceptions of the stuttering modification techniques “Cancellation” and “Pullout” compared with those of unmodified core behaviors. They videotaped a male speaker addressing the camera and producing a monologue about his family and social interests. He simulated repetitions, prolongations, and broken words. These fluency breaks were either modified or not. The modified speech sample was perceived as, for example, more anxious, afraid, nervous, unfriendly, and self-derogatory than the stuttering only speech sample.

Although the discussed studies provide important insights, they are limited in several ways. First, they focus on core behaviors and tell us little about associated behaviors and prolonged speech. Second, core behaviors were often simulated in a rather “standardized” way, in that repetitions and prolongations were produced “in an easy, rhythmical, and relaxed manner” (Panico et al., 2005, p. 69). Broken words (“blocks”) were mostly left out. The samples did not contain any associated facial or head movements “in order to eliminate this factor as a confounding variable” (Panico et al., 2005, p. 69). As a result, these studies were able to examine listener perceptions of “pure stuttering” but were likely to have limited external validity. Many people who stutter do not stutter in an easy, rhythmical, and relaxed manner, but have uncontrollable blocks that can be seen in their faces. Third, not only the simulated speech but also the settings of the samples were rather artificial. The participants watched a person producing a monologue or even reading a text and were then asked to rate his or her personal characteristics. In everyday life, however, people get to know a person who stutters in a dialogue, pursuing specific goals and having specific emotions and expectations. In these difficult social situations of everyday life, people who stutter can be expected to be perceived more negatively than in a monologue setting leading listeners to thoughts like “Oh, that's a stutterer in a therapy room. Stuttering is okay. I like him.” Again, the perfect “standardization” of the social setting may have resulted in poor external validity.

Thus, the purpose of the present study was to examine listener perceptions of different speech patterns often produced by people who stutter, for example core behaviors, interjected speech fragments, and prolonged speech. These different behaviors were simulated in a rather unstandardized manner. Furthermore, the presented video samples contained a specific social situation that was familiar to the participants so that they should have been able to put themselves in the position of the listener shown in the video clip. Therefore, participants did not have to make an abstract judgment about the person who stuttered, but were asked the probably easier question of how they would have perceived that person if they had been in that situation.

In a phenomenological analysis, Susca and Healey (2002) found that people listening to the speech of a person who stutters may think and draw conclusions about many different things. First, they make judgments about the speech itself, for example, regarding its rate, naturalness, or fluency. Second, they perceive their own reactions to the speech, for example their listener comfort. Third, they draw conclusions about the personality and competencies of the speaker, for instance intelligence or emotional stability.

The present paper focuses on listener perceptions of the third type. Therefore, recent studies concentrating on speech naturalness ratings (Kalinowski et al., 1994, Stuart and Kalinowski, 2004) or on listeners’ self-ratings when hearing stuttered speech (Guntupalli, Everhart, Kalinowski, Nanjundeswaran, & Saltuklaroglu, 2007) will not be discussed in detail here. Constructs like naturalness or listener comfort are important tools for treatment evaluation, but they tell us little about the processes of social anxiety and avoidance. A woman speaking with a French accent may be aware of sounding unnatural to her listeners, but usually neither feels anxious nor avoids the accent. Only if the woman believes that her accent makes her appear stupid, boring, arrogant, or communicatively incompetent to her listeners will she probably show reactions of anxiety and avoidance. It is this dimension of disrespect or disapproval in which the present work is interested.

Susca and Healey (2002) found many examples of those personality and competence conclusions, increasing with frequency of dysfluency. A person who stuttered was perceived as, for example, nervous, awkward, not believable, low in intelligence, giving odd descriptions, and having problems putting words and letters in the right order. The authors concluded that “listeners use the speech signal to make judgments of a speaker's personality, ability to think, and ability to use language for self-expression” (p. 157). Using quantitative designs, several studies (Evans et al., 2008, Panico et al., 2005, Susca and Healey, 2001) found that the more dysfluencies a male who stuttered produced, the less he was perceived as a competent speaker. However, differences between groups were often small and not always statistically significant.

These qualitative and quantitative results regarding core behaviors correspond nicely with well-established findings of stereotype research. MacKinnon et al. (2007) summarize available studies stating that persons who stutter are stereotyped as shy, insecure, reticent, guarded, avoidant, introverted, quiet, hesitant, self-derogatory, nervous, tense, and afraid. These judgments seem to be the result of experiencing “normal” dysfluencies; that is, people make judgments about people who stutter based on their own feelings when they themselves experience stuttering-like speaking moments (MacKinnon et al., 2007, White and Collins, 1984). Therefore, someone speaking very dysfluently is likely to be perceived as a very shy, nervous, self-derogatory person.

Both listener perception and stereotype studies suggest that, at least in Western cultures, stuttering is connected with two kinds of deficiencies. People who stutter are considered to be emotionally unstable (e.g., nervous, awkward, self-derogatory) and to be unable to communicate effectively in everyday life (e.g., not a competent speaker, giving odd descriptions, shy, quiet, introverted), two dimensions that are well known in social psychology (Jones and Pittman, 1982, Lopes et al., 2004). This interpretation is supported by several studies about stigmatizing (see Daniels & Gabel, 2004). In particular, Gabel, Blood, Tellis, and Althouse (2004) showed that stuttering negatively affects the perceptions of appropriate career choices for people who stutter. In a hypothetical scenario, university students advised a person who stuttered against becoming an attorney, judge, protestant minister, or psychologist and several other professions. The authors conclude that these are careers that require a great deal of communicative competence, assertiveness, and empathic listening ability.

In the present paper, the two capacities being questioned by listeners are called emotional competence and communicative competence. Emotional competence is a broad construct, including the abilities to express and perceive emotions appropriately, to have good self-esteem, to have good relationships with other people, and to be liked by other people (Petrides, Furnham, & Mavroveli, 2007). Communicative competence is the ability to adapt messages effectively and appropriately to the interaction context (Rickheit et al., 2008, Spitzberg and Cupach, 1984; see also Blood et al., 2001). A person who is perceived as communicatively incompetent is seen as having problems in making his or her point clear, making arrangements, and avoiding misunderstandings.

How much a person who stutters is perceived as emotionally and communicatively incompetent may depend on the chosen way of speaking. Manning et al. (1999) found that people using stuttering modification techniques were perceived as, for example, more anxious, afraid, nervous, unfriendly, and self-derogatory than people producing core behaviors only. The present study examines four different ways of speaking, labeled as: (1) stuttered speech, (2) stuttered/hesitant speech, (3) hesitant speech, and (4) prolonged speech.

Stuttered speech is a way of speaking that contains core behaviors only.

Hesitant speech contains no core behaviors, but a special kind of associated behaviors, namely, interjections (starters, fillers), revisions, incomplete phrases, and pauses, occurring when the speaker seeks to avoid core behaviors (Guitar, 2006). Bloodstein and Bernstein Ratner (2008) call these symptoms “interjected speech fragments” (p. 17), while others may call them normal dysfluencies. However, to avoid misunderstanding, the new term “hesitant speech” is preferred here. Hesitant speech contains more and longer interjections, revisions, incomplete phrases, and pauses than the speech of most normally fluent people. It is, like stuttered speech, a form of coping with the feeling of stuttering, or, in a word, a form of stuttering. Manning (2001) argued that someone producing hesitant speech can be “thought of as someone who is introverted, shy, lazy, or at the very least, somewhat peculiar” (p. 123). The hesitant and inappropriate communication behaviors implied in avoidance strategies can make the person who stutters look more emotionally disturbed and disagreeable than before (Petrunik & Shearing, 1983). However, a person producing hesitant speech can hardly be seen as a “stutterer” (i.e., having a chronic communication disorder) but rather as someone having a temporary problem (e.g., being tired, drunken, bored, annoyed, nervous, anxious, or depressive). That is, while stuttered speech should be associated with communication problems directly, hesitant speech should be taken as an indicator of personality problems. As a result, it can be hypothesized that a person using hesitant speech should be regarded as more communicatively competent but less emotionally competent than a person using stuttered speech.

Stuttered/hesitant speech is a speech that contains both core behaviors and interjected speech fragments. Most people who stutter are not able to or do not want to avoid every moment of stuttered speech. Therefore, a combination of stuttered speech and hesitant speech was examined as well, expecting low levels of both perceived emotional and communicative competence.

Prolonged speech is a speech pattern that is learned in fluency-shaping therapies (e.g., O’Brian et al., 2003a, Webster, 1974). The main characteristics of prolonged speech are syllable prolongations, gentle voice onsets, smooth sound transitions, and light articulatory contacts. Prolonged speech sounds less natural to listeners than normal fluent speech (e.g., Onslow and Packman, 1997, Stuart and Kalinowski, 2004). While there are several studies about speech naturalness, little is known about what personality attributes or competencies are attributed to a person using prolonged speech. However, it can be expected that people producing prolonged speech are regarded as more emotionally incompetent than people producing stuttered speech, because, like interjections and incomplete phrases, a rather monotonous and artificial way of speaking seems likely to prevent the speaker from expressing his or her personal needs and wishes (Kamhi, 2003). Furthermore, as mentioned above, Manning et al. (1999) have found this effect with another kind of post-treatment speech, namely stuttering modification.

There may be several variables that mediate the relationship between the chosen way of speaking and listener perceptions. One of them may be the implicit theory of the perceived speaking problems held by the listener. The concept of implicit theory, as developed by Dweck, 1999, Dweck and Leggett, 1988), differentiates between an incremental theory and an entity theory. People holding an incremental theory believe that personal attributes (e.g., intelligence) can be changed, whereas those holding an entity view believe in the idea of stable personal attributes. It could be shown that implicit theories influence a broad variety of psychological processes, including perception, attribution, and action (Molden & Dweck, 2006). In the context of stuttering, an entity theory of stuttering can be defined as holding that stuttering is a genetically based physical defect resulting in a chronic speech disorder, and an incremental theory as holding that stuttering is the temporary result of psychological factors that can be changed. Most recently, Boyle, Blood and Blood (2009) induced implicit theories (or causality attributions, as they labeled it) experimentally using a vignette research methodology. Incremental participants (who read that stuttering was due to psychological factors) made more negative judgments and higher social distance ratings towards the target person than entity participants (who read that stuttering was due to genetic factors). However, parallel to the considerations made concerning hesitant speech, it can be expected that listeners holding an entity view of perceived speaking problems will tend to regard the speaker as communicatively incompetent, whereas people holding an incremental view will tend to regard the speaker as emotionally incompetent.

In summary, the present study examines listeners’ perceived emotional competence, perceived communicative competence, and implicit theory, of stuttered speech, hesitant speech, stuttered/hesitant speech, and prolonged speech. That is, four dysfluent ways of speaking are examined regarding three questions: How does a listener perceive a person using this way of speaking? Are there differences regarding the dimensions of evaluation a listener uses? Do the perceptions depend on the listeners’ causal attribution of the way of speaking?

Whereas many studies of listener perceptions include a fluent control condition, the present study does not. Like Manning et al. (1999), the present study is not interested in comparing perceptions of dysfluencies to those of fluency, but to compare perceptions of different kinds of dysfluencies. These are different questions, resulting in different methodologies. While numerous stereotype studies could show that people who stutter are associated with negative personality traits, the present research seeks to find out which of their behaviors may be responsible for those stereotypes. Therefore, this study is not about the question if people who stutter are perceived more negatively than people who do not stutter. Rather, it seeks to inform people who stutter of how they are perceived depending on the way of speaking they use.