2003 IFA Congress: Montreal, Canada

Verbal Behavior of Listeners Interacting with a Person Who Stutters

Debora Freud, Ronit Sharir and Ruth Ezrati-Vinacour
Department of Communication Disorders, Sackler Faculty of Medicine, Tel Aviv University, Israel.


The purpose of this study is to investigate (a) whether listeners speak faster to adult PWS than to fluent speakers, and (b) whether listeners interrupt, reinforce and complete PWS’ speech more frequently than fluent speakers’ speech. Ten adult listeners spoke once with an adult PWS and once with a fluent speaker. Analysis of speech samples demonstrated that the listeners spoke faster to the PWS than to the fluent speaker. No significant differences were found between listeners’ interruptions, sentence completion and reinforcement to the two speakers. However, listeners performed more interruptions and sentence completions while the PWS was stuttering than while speaking fluently


  1. Introduction
A conversation is a complex social interaction, in which changes in the paralinguistic behavior of one participant may influence the other.participant’s paralinguistic behavior. The term “Recipient Design” refers to the fact that speech produced during conversation by one of the participants is designed in a way that expresses sensibility to the other participant’s true or assumed needs, abilities, background and knowledge (Have, 1999; Newman & Smith, 1989). Several studies show that when people talk to different kinds of populations, their characteristics of speech change. People change their speech so that it would fit the true or assumed features of the people they talk to (Beisecker & Thompson, 1995; Depaulo & Coleman, 1986; Fernald & Mazzie, 1991; Harwood et al., 1995).

When talking to people who stutter (PWS) we often hear about the way they feel about their conversational partners; PWS often feel that their listeners speak to them in a manner that is different from the one usually used with fluent speakers. Quite frequently PWS feel they are being laughed at or thought to be unintelligent (Bloodstein, 1975). Starkweather (1987) noted that the cosmetic features of stuttering may have an impact on listeners, apart from the stuttered speech itself.

The approach to stuttering as a communication disorder and not merely a specific speech disorder led us to investigate the way stuttering influences the communicational interaction.

Studies have examined the way listeners’ behavior affects stuttering (Meyers & Freeman, l985a), and the way stuttering influences listeners’ behavior (Kelly & Conture, 1992), focusing primarily on interactions between children who stutter and their parents. To the best of our knowledge, so far no study has dealt with the behavior of adult fluent listeners while speaking to adult PWS.

The purpose of this study was to investigate the influence of stuttering in adults on the paralinguistic behavior of adult listeners, by examining the following: 1. Whether adult listeners speak at a different speech rate to PWS than to fluent speakers. 2. Whether listeners' turn taking behavior (including interruptions, sentence or word completions and reinforcers) with PWS differ from turn taking with fluent speakers.

  1. Method

Listeners’ group included 10 young women (mean age=25). As far as we know, none of the listeners had previous experience with PWS. Speakers were 2 adult men, matched by age, education, family status, employment, and appearance. One speaker had fluent speech and the other speaker had stuttered speech. Stuttering severity, as assessed by three speech therapists was rated as severe.


Each listener spoke twice- once with the PWS and once with the fluent speaker, at a randomized order. Listeners were instructed to speak with speakers for 10 minutes about a specific topic. The topic varied between listeners but was the same for each listener with both speakers. Prior to interactions, listeners were informed about the speakers’ background, age, family, etc. No information was given ahead of time concerning the speaker’s speech nor was the purpose of this study revealed. Conversations were videotaped and audio-taped. A total of 20 interactions were recorded.

Data coding

The recorded data were carefully transcribed and divided into turns. A conversational turn is defined as a segment of speech produced by a speaker, which is bound in its beginning and its end by the other speaker’s speech, and includes all the utterances a speaker produces until the other speaker begins to talk again (Kelly & Conture, 1992). Turns were classified into disfluent and fluent categories. Disfluent turns were categorized as either stuttering type disfluencies or normal-type disfluencies. The stuttered turns included part word repetitions, prolongations, and tense pauses (Meyers & Freeman, 1985b). The normal-type disfluent turns included whole word or phrase repetitions, revisions, incomplete phrases and interjections (Meyers & Freeman, 1985b). Data coding was done by two judges (speech clinicians).

To obtain the average speech rate (syllables per second), the number of syllables in each speaking turn was counted and divided by the total number of seconds elapsed for the turns.

The term “turn taking behavior” included the following variables (Wiemann & Knapp, 1975)

Interruptions: Simultaneous speaking of the listener and the speaker, as an attempt of the listener to assume the speaking role before it has been relinquished by the current speaker.

Sentence/word completions: Utterances produced by the listener in order to complete the word or sentence uttered by the speaker.

Reinforcers: Words produced by the listener that provide feedback to the speaker (e.g., “um- hm”, “yes” and “right”).

These types of behavior were classified into ones made during fluent turns and ones during disfluent turns.

During data analysis we have found that speakers differed in the number of short and long turns they have produced. This difference was not statistically tested, but was noticeably observed and calculated. A short turn was defined as a turn that contained two words at most. A long turn was defined as one that contained at least 3 words. 43% of the PWS’s turns were short, and 57% were long, whereas 23% of the turns of the fluent speaker were short and 77% long. Therefore, in order to assure that the statistic analysis would not be influenced by the length of turn variable, and focus on the speech rate analysis, it was decided to divide speaking turns into the following categories:
  1. short fluent turns
  2. short disfluent turns
  3. long fluent turns
  4. long disfluent turns.
Since different speakers would produce different number of turns, all analyses were conducted using proportional measures (%).


To obtain interjudge reliability, 50% of interactions were randomly selected and coded independently by the two speech clinicians. The results of each of the analyses performed by the speech clinicians were compared using Pearson Correlation Coefficient. Reliabilty for speech rate was r=0.84; The mean reliabilty for overall turn taking behaviors was r=0.88 (range r=0.707- r=O.98; p< 0.05).

  1. Results and Discussion
Listeners’ speech rate

In order to compare statistically listeners’ speech rates, Wilcoxon Signed Rank Tests were performed. Statistically significant differences (z=2.65, N=l0; sum of negative ranks:-1, sum of positive ranks=54; p<0.0l) were found between listeners’ speech rates with listeners’ speech rate being faster when speaking to the PWS than with the fluent speaker (see Table 1)


Table 1. Listeners’ mean speech rates ( and SDs) during conversation with the PWS and the fluent speaker

This is in accordance with Meyers and Freeman (1985a), who found that mothers of stuttering children and nonfarniliar women spoke faster to stuttering children than to fluent children. A number of explanations for this finding are possible. First, listeners may feel discomfort when interacting with a stutter. This might cause the listener to increase his speech rate (Starkweather, 1987). Second, stuttering may disrupt the pace of the ongoing conversation, thus prompting the listener to increase his speech rate in hope to increase the PWS’s speech rate. Third, since PWS’ speech time tends to be longer than that of fluent speakers, it is possible that listeners increased their speech rate in order to compensate for the short time left for them to speak (Meyers & Freeman, 1985a).

According to Sheehan (1975) and Meyers and Freeman (1985a), speech rate is an important variable in conversation, especially when interacting with a PWS. A relatively fast speech rate towards a PWS might cause time pressure, thus make the conversation harder for the PWS and therefore increase his stuttering.

Turn taking behavior performed by listeners towards speakers

Wilcoxon Signed Rank Tests showed no significant differences between the proportional number of all turn taking behaviors performed by listeners towards the PWS as compared to those performed towards the fluent speaker (see Table 2).


Table 2. Means (%) and SDs of listener ‘s turn taking behaviors during conversations with the fluent speaker vs. conversations with the PWS.

These findings are in contradiction to our assumption, which was based on PWS’s reports of being treated differently during conversation, specifying these investigated variables, and also considering the various investigations that demonstrate misjudgment of PWS’s skills and negative attitudes (McKinnon et al., 1986; Patterson & Pring, 1991; Turnbauh et al., 1981). According to Wiemman and Knapp (1975), almost any verbal interaction is based on a turn taking mechanism that includes interruptions, completions and reinforcers, and therefore these were also performed towards the PWS. Thus, results may suggest, that PWS expect an unnatural and artificial structure of conversation which includes a minimum of turn-taking behaviors exhibited by their listeners.

Turn taking behavior performed by listeners during fluent and disfluent turns of each speaker

Examination of the interactions with the fluent speaker (by Wilcoxon Signed Rank Tests) revealed no significant differences in the proportional number of turn taking behaviors performed during fluent or disfluent turns (see Table 3).


Table 3. Means (%) and SDs of listeners ‘ turn-taking behaviors performed towards fluent speakers disfluent vs. fluent speaker is fluent turns.

Interestingly, Wilcoxon Signed Rank Tests showed significant differences between the overall proportional number of turn taking behaviors performed towards the PWS's stuttered and fluent turns. Specifically, when keeping aside the turn length variable, examination showed that the lack of fluency in the stutter's speech turns did affect listeners’ verbal behavior, which was characterized by a significant increase in the overall number of turn taking behavior types (z=2.039, N=l0; sum of negative ranks:-48, sum of positive ranks=7;_p<0.05) and similarly an increase of sentence/word completions (z=2.446, N=10; sum of negative ranks=-52, sum of positive ranl<s=3; p<0.05). No differences were found betweeen the number of reinforcers and interruptions performed during fluent and stuttered speech turns. Results are shown in Table 4.


Table 4. Means (%) and SDS of listeners’ turn-taking behaviors performed towards PWS 's disfluent vs. PWS’s fluent turns.

Similar findings were reported by Meyers and Freeman (l985b), in their research on the verbal behavior of mothers interacting with children who stutter and children who speak fluently. Several suggestions may explain this finding. We find these types of behavior to be more frequent during the stuttering, a type of disfluency which was unfamiliar and unusual to them, in comparison to the disfluencies performed by the fluent speaker. Naturally the difference between the fluent turns and stuttered turns of the PWS were much greater and noticeable than the slight differences between the fluent and disfluent turns of the fluent speaker. Therefore, no special attention was directed towards the disfluent turns of the fluent speaker, as opposed to the stuttered turns that did have an impact on the listeners. It is possible that this unfamiliar disfluency was misinterpreted as a sign of difficulty in pulling out a word or expressing an idea, and thus listeners performed assistive behavior that included interruptions and completing utterances (Meyeres & Freeman, 1985b). In addition, listeners may have misinterpreted these unfamiliar stuttered disfluencies as signs of giving up his turn (Meyers & Freeman, 1985b; Duncan, 1972). Finally, listeners may have been impatient when stuttering occurred and therefore they performed more overall turntaking behaviors during the stuttered turns.

  1. Conclusions
Comparison of the two types of conversations- one with the fluent speaker and one with the PWS- revealed a single difference in listeners’ speech rate. Only when comparing listeners’ turn taking behaviors during stuttered and fluent turns of the PWS we have found differences in several variables. The finding that listeners increase their speech rate and tend to perform more turn taking variables during the moments. of stuttering, can explain PWS’s feelings about their listeners’ verbal behavior.

Our findings suggest including a discussion about the investigated variables of a verbal interaction during treatment of PWS in regard to the following aspects. First, discussing the common elements of a verbal interaction and the conflict of the wish for natural conversation versus adjusted or artificial conversation with PWS. Second, demonstrating and practicing strategies for coping with listeners’ increased speech rate. Speech clinicians may suggest PWS to ask their partners to speak slower, and as Gregory (1991) suggested also teach PWS some techniques for coping with the time pressure.

Beisecker, A.E. & Thompson, T.L.( 1995). The elderly patient- physician interaction. In: J .F. Nussbaum and J. Cupland (Eds.), Handbook of communication and aging research. NJ: Lawrence Erlbaum Associates, Inc.
Bloodstein, O. (1975). Stuttering as tension and fragmentation. In J . Eisenson (Ed.), Stuttering: A second symposium. NY: Harper & Row.
DePaulo, B.M.& Coleman, J .M. (1986). Talking to children, foreigners and retarded adults. Journal of Personality and Social Psychology, 51, 945-959.
Duncan, S. (1972). Some signals and rules for taking speaking turns in conversations. Journal of Personality and Ssocial Psychology 23. 283-292.
Fernald, A. & Mazzie, C. (1991). Prosodyiand focus in speech to infants and adults, Developmental Psychology, 27, 209-221.
Gregory, H. (1991). Stuttering therapy: Workshop for specialists. Northwestern University: The Speech Foundation of America.
Harwood, J ., Giles, H. & Ryan, E.B.(1995). Aging, communication and Intergroup Theory: Social Identity and intergenerational communication. In: J .F. Nussbaum and J . Cupland (Eds.), Handbook of Communication and Aging Research.. NJ: Lawrence Erlbaum Associates, Inc. Section 5. Language, Speech and Discourse 30]
Have, P.T. (1999). Doing conversational analysis: A Practical Guide, Surrey (Great Britain): Sage Pub. Ltd. Kelly, E.M. & Conture, E.G. (1992). Speaking rates, response time latency, and interrupting behaviors of young stutterers, nonstutterers, and their mothers. Journal of Speech and Hearing Research, 35,_1256-1267. A
McKinnon, S.L, Hess, C.W. & Landry, R.G. (1986). Reaction of college students to speech disorders. Journal of Communication disorders, 19, 75-82.
Meyers, S.C. & Freeman, F.J.( 1985a). Mother and child speech rates as a variable in stuttering and disfluency. Journal of Speech and Hearing Research, 28, --36-444.
Meyers, S.C. & Freeman, F.J. (1985b). Interruptions as a variable in stuttering and disfluency. Journal of Speech and Hearing Research, 28, 428-435.
Newmann, L.L & Smith, A.B. (1989). Some effects of variation in response time latency on speech rate, interruptions and fluency in children’s speech. Journal of Speech and Hearing Research, 32, 635-64.
Patterson, J. & Pring T. (1991). Listeners attitudes to stuttering speakers. Journal of Fluency Disorders, 16, 201-205.
Sheehan, J .G. (1975). Conflict theory and avoidance-reduction therapy. In J. Eisenson (Ed.), Stuttering: A second symposium. NY: Harper & Row.
Starkweather, C.W.(l987). Fluency and stuttering, NJ: Prentice Hall, Inc.
Turnbaugh, K., Guitar, B. & Hoffman, P.(1981) The attribution of personality traits: the stutterer and nonstutterer. Journal of Speech and Hearing Research, 24, 288-291.
Wiemann, J.M. & Knapp, M.L.(1975). Turn taking in conversations. Journal of Communication, 25, 75-92.


In preparation for the 2018 World Congress the IFA is implementing Japanese translations of some pages on the site. Choosing Japanese below to see these translations.

Not all pages are translated, but you can use Google translate to see a machine translation using the switch below

Google Translate

Follow the Joint World Congress