The significance of pauses in spontaneous speech

Studies of filled and silent pauses performed in the last two decades are reviewed in order to determine the significance of pauses for the speaker. Following a brief history, the theoretical implications of pause location are examined and the relevant studies summarized. In addition, the functional significance of pauses is considered in terms of cognitive, affective-state, and social interaction variables.     

Questions and the allocation,construction, and timing of turns in child discourse

The role of questions (i.e., interrogative forms with illocutionary force of questioning) in turn-taking was investigated. Turn-taking has three components: speaker selection, turn construction, and timing of turn transfer. Questions explicitly, via the verbal code, provide a clearance signal and a requested topic for the response, whereas assertions do not. Consequently, questions, in general, but not all linguistically defined sub-types of questions, contrast with assertions on the first two turn-taking components. It was shown that the probability of a turn transfer, of a relevant response, and of a turn transfer and relevant response is higher following questions, in general, than assertions. Questions also transmit an earlier speaker-selection signal (initial or terminal clausal position) whereas in the case of assertions speaker selection must be managed following clause completion. Questions do not seem to contrast with assertions on the third component as a result of this earlier clearance signal. Keeping the relevance of an utterance to prior text and the content of the response constant, switching pauses were about the same following both questions and assertions.This research was supported from funds of the Biomedical Research Support Grant, 5SO7RR07041-12 to The Johns Hopkins University and a grant from the Spencer Foundation to the second author.     

Analyzing nonverbal listener responses using parallel recordings of multiple listeners

In this paper, we study nonverbal listener responses on a corpus with multiple parallel recorded listeners. These listeners were meant to believe that they were the sole listener, while in fact there were three persons listening to the same speaker. The speaker could only see one of the listeners. We analyze the impact of the particular setup of the corpus on the behavior and perception of the two types of listeners: the listeners that could be seen by the speaker and the listeners that could not be seen. Furthermore, we compare the nonverbal listening behaviors of these three listeners to each other with regard to timing and form. We correlate these behaviors with behaviors of the speaker, like pauses and whether the speaker is looking at the listeners or not.     

Analyzing nonverbal listener responses using parallel recordings of multiple listeners

In this paper, we study nonverbal listener responses on a corpus with multiple parallel recorded listeners. These listeners were meant to believe that they were the sole listener, while in fact there were three persons listening to the same speaker. The speaker could only see one of the listeners. We analyze the impact of the particular setup of the corpus on the behavior and perception of the two types of listeners: the listeners that could be seen by the speaker and the listeners that could not be seen. Furthermore, we compare the nonverbal listening behaviors of these three listeners to each other with regard to timing and form. We correlate these behaviors with behaviors of the speaker, like pauses and whether the speaker is looking at the listeners or not.     

Thee,uhh disfluency effect in preschoolers: A cue to discourse status

Speech disfluencies, such as filled pauses (ummm, uhhh), are increasingly recognized as an informative element of the speech stream. Here, we examined whether 2‐ and 3‐year‐olds expected that the presence of filled pause would signal reference to objects that are new to a discourse. Children viewed pairs of familiar objects on a screen and heard a speaker refer to one of the objects twice in succession. Next, children heard a critical utterance and were asked to look and point at either the discourse‐given (i.e., previously mentioned) or discourse‐new (i.e., previously unmentioned) object using a fluent (‘Look at the ball!’) or disfluent (‘Look at thee uh ball!’) expression. The results indicated that 3‐year‐old children, but not 2‐year‐old children, initially expected the speaker to continue to refer to given information in the critical utterance. Upon hearing a filled pause, however, both 2‐ and 3‐year‐old children's looking patterns reflected increased looks to discourse‐new objects, although the timing of the effect differed between the age groups. Together, these findings demonstrate that young children have an emerging understanding of the role of filled pauses in speech.     

The use of prosody in highlighting alterations in repairs from unrestricted speech

A speaker has several ways in which he or she may highlight the fact that an error or imprecision of speech has been made and subsequently altered. The three principal ones are by signalling through the structure of the speech that surrounds the error (the repair-syntax), by the use of prosody, and through the semantic content. The role of prosody in the correction process is investigated in the current studies. Analysis of the prosody of a number of errors and their alterations drawn from unrestricted speech are reported. The analysis shows that pauses occur at the moment of interruption and that an increase in stress occurs at the start of the alteration. Pauses could indicate the moment of interruption, and stress could highlight what has been altered. Two sets of perceptual experiments were carried out to assess whether these cues are salient for listeners who hear constructions containing an error and its alteration. Two paradigms were employed in each set of experiments: (1) direct judgement about the comprehensibility of sentences containing errors and alterations, and (2) repeating a message that had an error and alteration without the error. The effects of stress and pauses on (Experiments 1A and 1B) or pauses around (Experiment 2A and 2B) the alterations were assessed. In the first set of experiments it was shown that pauses and stress help listeners process repairs. When a word is spoken in error, the speaker may repeat a section of speech immediately preceding the alteration and/or a section immediately following that word. Inclusion of these repeated sections allows assessment of whether pauses signal where the interruption occurred. The second experiment shows that the placing of the pause before the retrace, rather than at other locations, indicates to listeners where the repair starts.     

Is the fluency of language outputs related to individual differences in intelligence and executive function?

There has been little research on the fluency of language production and individual difference variables, such as intelligence and executive function. In this study, we report data from 106 participants who completed a battery of standardized cognitive tasks and a sentence production task. For the sentence production task, participants were presented with two objects and a verb and their task was to formulate a sentence. Four types of disfluency were examined: filled pauses (e.g. uh, um), unfilled pauses, repetitions, and repairs. Repetitions occur when the speaker suspends articulation and then repeats the previous word/phrase, and repairs occur when the speaker suspends articulation and then starts over with a different word/phrase. Hierarchical structural equation modeling revealed a significant relationship between repair disfluencies and inhibition. Conclusions focus on the role of individual differences in cognitive ability and their role in models and theories of language production.     

THE VOCABULARIES OF ACADEMIA

Abstract— It has been demonstrated that humanists are far more likely to use filled pauses ("uh,""ah," or "um") during their lectures than are social or natural scientists This finding has been interpreted in terms of the hypothesis that filled pauses indicate time out while the speaker searches for the next word or phrase Based on the assumption that the more options at a choice point, the more likely a speaker will say "uh," it is hypothesized that the humanities are characterized by richer vocabularies (i e, more synonyms) than are the sciences An analysis of the number of different words used in lectures and in professional publications indicates that this is indeed the case Scientists consistently use fewer different words than do humanists Further, the number of different words correlates positively with the frequency of saying "uh" during lectures These findings are not restricted to academics, for in newspaper accounts, journalists use fewer different words in stories about science than in stories about the arts     

Specific effects of the Trier Social Stress Test on speech fluency in young and older adults

The notion that speech becomes less fluent during stressful speaking conditions has received little empirical test, and no research has tested this relationship in older adult participants. We analyzed speeches produced during the Trier Social Stress Test (TSST) or during a less stressful placebo (pTSST) version of the task. We measured young and older adults’ speech fillers (e.g., um), unfilled pauses (at least 1 s in duration), and other disfluencies (e.g., repetitions, repairs). Neither young nor older adult participants rated themselves as having greater stress in the TSST than pTSST condition, but behavioral effects were obtained. Participants in the TSST condition produced more mid-phrase speech fillers and unfilled pauses than participants in the pTSST condition. Young adults produced more unfilled pauses than older adults overall, and older adults produced more mid-phrase fillers than young adults. Critically, age group interacted with experimental condition, such that older speakers produced disproportionately more mid-phrase fillers than young adults in the TSST compared to the pTSST condition. In sum, the negative effects of the TSST on fluency were generally similar across age, but this specific age-related increase in mid-phrase fillers indicates that older adults’ word retrieval may have been particularly negatively affected. Findings are generally consistent with previous research and add to understanding of how factors internal to the speaker (i.e., demographic, personality, and cognitive variables) and factors external to the speaker (i.e., variables regarding the situation, context, or content of speech) combine to affect speech fluency.     

The voice of confidence: Paralinguistic cues and audience evaluation

A standard speaker read linguistically confident and doubtful texts in a confident or doubtful voice. A computer-based acoustic analysis of the four tapes showed that paralinguistic confidence was expressed by increased loudness of voice, rapid rate of speech, and infrequent, short pauses. Under some conditions, higher pitch levels and greater pitch and energy fluctuations in the voice were related to paralinguistic confidence. In a 2 × 2 design, observers perceived and used these cues to attribute confidence and related personality traits to the speaker. Both text and voice cues are related to confidence ratings; in addition, the two types of cue are related to differing personality attributes.     

Paralinguistic correlates of conceptual structure

How is conceptual knowledge transmitted during conversation? When a speaker refers to an object, the name that the speaker chooses conveys information about categoryidentity. In addition, I propose that a speaker’s confidence in a classification can convey information about categorystructure. Because atypical instances of a category are more difficult to classify than typical instances, when speakers refer to these instances their lack of confidence will manifest itself “paralinguistically”—that is, in the form of hesitations, filled pauses, or rising prosody. These features can help listeners learn by enabling them to differentiate good from bad examples of a category. So that this hypothesis could be evaluated, in a category learning experiment participants learned a set of novel colors from a speaker. When the speaker’s paralinguistically expressed confidence was consistent with the underlying category structure, learners acquired the categories more rapidly and showed better category differentiation from the earliest moments of learning. These findings have important implications for theories of conversational coordination and language learning.     

Talking to children, foreigners, and retarded adults

Speech addressed to different categories of listeners was examined in a study in which undergraduate women taught a block design task to either a 6-year-old child, a retarded adult, a peer who spoke English as a second language (foreigner), or a peer who was an unimpaired native speaker of English. The speech addressed to children differed from the speech addressed to native adults along every major dimension that emerged in this study: It was clearer, simpler, and more attention maintaining, and it included longer pauses. Speech addressed to retarded adults was similar in numerous ways to the speech addressed to 6-year-olds; in some ways (e.g., repetitiveness), it was even more babyish. However, speech to the retarded adults did differ in timing from the other styles of speaking in that it included fewer and somewhat shorter pauses. Speech addressed to foreigners was more repetitive than speech addressed to native speakers, but in all other ways it was very similar. There was some evidence that speakers fine-tuned their communications to the level of cognitive and linguistic sophistication of their particular listener; for example, speakers addressing the more sophisticated foreigners (relative to those addressing the less sophisticated foreigners) used speech that included fewer devices for clarifying, simplifying, and maintaining the listeners' attention. We discuss the hypothesis that baby talk (the speech addressed to children) is a prototypical special speech register from which other special registers are derived.     

An experimental investigation of the modifiability of the temporal structure of spontaneous speech

This study attempted to test the hypothesis that the temporal structure of spontaneous speech is modifiable by reinforcing and punishing pauses, of a certain duration, in an operant conditioning situation. Pause rate was significantly affected by these contingencies: moreover, rate of change was rapid, indicating a prepared association between pausing and such contingencies. This study also attempted to test the hypothesis that there is a class of noncognitive pauses in monologue by punishing UPs to determine if UPs can be eliminated without affecting speech content. Although this manipulation did lead to a decline in pause rate, a significant increase in the amount of filled hesitation, particularly in repetition, resulted. This suggests that the overall amount of hesitation is fixed by the cognitive demands of the task but that a speaker is able to adapt to different interactional contexts by varying the category of hesitation used for cognitive planning.     

Toddlers use speech disfluencies to predict speakers' referential intentions

The ability to infer the referential intentions of speakers is a crucial part of learning a language. Previous research has uncovered various contextual and social cues that children may use to do this. Here we provide the first evidence that children also use speech disfluencies to infer speaker intention. Disfluencies (e.g. filled pauses 'uh' and 'um') occur in predictable locations, such as before infrequent or discourse-new words. We conducted an eye-tracking study to investigate whether young children can make use of this distributional information in order to predict a speaker's intended referent. Our results reveal that young children (ages 2;4 to 2;8) reliably attend to speech disfluencies early in lexical development and are able to use disfluencies in online comprehension to infer speaker intention in advance of object labeling. Our results from two groups of younger children (ages 1;8 to 2;2 and 1;4 to 1;8) suggest that this ability emerges around age 2.     

Vocal congruence in mother-infant play

Turn-taking is the fundamental temporal structure of adult dialogue. This structure defines two types of joint silence: intrapersonal pause (silence bounded by the vocalizations of a single speaker) and switching pause (silence bounded by the vocalizations of different speakers). Switching pauses mark the boundaries of the turn exchange. In adult conversation the mean durations of both types of pause are characteristically matched between partners. This matching, termed vocal congruence, occurs developmentally earlier in the case of switching pauses. We hypothesized and confirmed that mothers and infants match switching pauses but not intrapersonal pauses at 4 months, even though the infants' vocalizations are prelinguistic. Second, since there are known affective correlates of vocal congruence in adult conversation, we hypothesized a similar affective correlate for mother-infant vocal congruence. We found, for the intrapersonal pause only, that the degree of matching within a dyad correlated with infant affective engagement. We conclude, from switching pause congruence, that a turntaking dialogic structure is being regulated in the mother-infant pair at 4 months in the same way as seen in adult conversation. Thus, both the temporal structure of adult dialogue and its affective correlate are prelinguistic.Supported in part by NIMH grant No. 41675.     

Speaker variations influence speechreading speed for dynamic faces

We investigated the influence of task-irrelevant speaker variations on speechreading performance. In three experiments with video digitised faces presented either in dynamic, static-sequential, or static mode, participants performed speeded classifications on vowel utterances (German vowels /u/ and /i/). A Garner interference paradigm was used, in which speaker identity was task-irrelevant but could be either correlated, constant, or orthogonal to the vowel uttered. Reaction times for facial speech classifications were slowed by task-irrelevant speaker variations for dynamic stimuli. The results are discussed with reference to distributed models of face perception (Haxby et al, 2000 Trends in Cognitive Sciences 4 223-233) and the relevance of both dynamic information and speaker characteristics for speechreading.     

Perceptions of unresponsive others: Attributions,attraction, understandability,and memory of their utterances

Analyses of the rules governing conversation have frequently pointed to the existence of a rule dictating that responses to others' communications should be relevant. Unfortunately, there have been few systematic theoretical analyses of the consequences of violations of the response relevance rule. The present research was designed to extend our previous analyses of the consequences of responsiveness in dyadic interaction to examine in detail the effects of irrelevant response on processing and retention of response content, and attributions concerning the speaker. It was argued that irrelevant responses will be processed and retained more poorly, because the preceding utterance will provide a less adequate context for their interpretation. In addition, it was argued that unresponsiveness will generally be attributed to the lack of one or more of four factors previously suggested to facilitate responsiveness: attention to the other, understanding of the preceding communication, adequate response repertoire, or motivation to be responsive—along with situational (e.g., distraction) or dispositional (e.g., incompetence) reasons for their absence. These processes were examined in the context of a political debate, in which the relevance of candidates' answers to the questions was varied. As expected, response relevance was shown to facilitate processing (as indexed by ratings of clarity and organization) and retention (both recall and recognition memory) of response content. Further, the unresponsive speaker was perceived to have less clearly understood the questions asked, to possess less knowledge of the facts and understanding of the issues concerning the questions raised, and to be less motivated to discuss the issues raised (as opposed to facilitating his self-presentational goals). Moreover, those attributions were accompanied by attributions of dispositional incompetence. Finally, the unresponsive candidate was perceived as less attractive than the responsive candidate.     

The effects of self-disclosure and non self-disclosure of stuttering on listeners' perceptions of a person who stutters

The aim of this study was to examine listener perceptions of an adult male person who stutters (PWS) who did or did not disclose his stuttering. Ninety adults who do not stutter individually viewed one of three videotaped monologues produced by a male speaker with severe stuttering. In one monologue, 30 listeners heard the speaker disclose stuttering at the beginning and in another monologue, 30 listeners heard the speaker disclose stuttering at the end. A third group of 30 listeners viewed a monologue where no disclosure of stuttering occurred. After listeners viewed a monologue, they were asked to rate a set of six Likert scale statements and answer three open-ended questions. The results showed that only one of six Likert statements was significantly different across the three conditions. The only statement that was different was that the speaker was perceived to be significantly more friendly when disclosing stuttering at the end of the monologue than when not disclosing stuttering. There were no significant differences between the percentage of positive and negative comments made by listeners across the three conditions. Listeners' comments to each open-ended question showed they were comfortable listening to stuttering with or without disclosure and slightly more than half of the listeners believed their perceptions of the speaker did not change when he disclosed stuttering. The results also showed that the speaker who disclosed stuttering at the beginning of the monologue received significantly more positive listener comments than when he disclosed stuttering at the end of the monologue. Results are discussed relative to comparisons with the study, the clinical relevance of acknowledging stuttering as a component of treatment, and future research on the self-disclosure of stuttering. Educational objectives: The reader will be able to: (1) describe how different groups of listeners perceive and respond to two conditions of self-disclosure of stuttering and one condition involving non self-disclosure of stuttering; (2) summarize the range of listener responses to and benefits of self-disclosure of stuttering; and (3) describe the value of self-disclosure of stuttering for the listener and the speaker.