The phenomena of disfluency clusters have been examined in the speech of children who stutter (CWS) and children who do not stutter (CWNS). Little is known about disfluency clusters in the adult population. The purpose of this study was to examine the characteristics of disfluency clusters in adults who stutter (AWS). The participants were ten AWS ranging in age from 18 to 60 (mean age = 35), with a stuttering severity of 9 to 30% (mean = 19%). Each participant provided a conversational speech sample of at least 300 words. Analysis focused on disfluency type, utterance length, speaking rate, and perceptual measures. Findings indicated that utterances containing disfluency clusters were significantly longer than fluent utterances and the speaking rate of fluent utterances was found to be significantly faster than that of disfluent utterances. Collectively the results appear to support a linguistic interpretation of disfluency clusters. The clinical implications of the findings are discussed.

Hesitations in speech marked by pauses, fillers such as er, and prolongations of words are remarkably common in most spontaneous speech. Experimental evidence indicates that they affect both the processing of speech and the lasting representation of the spoken material. One theory as to the mechanisms that underlie these effects is that filled pauses heighten listeners' attention to upcoming speech. For example, in the utterance: (1) She hated the CD, but then she's never liked my taste in er music The hesitation marked by the filler er would heighten listeners' attention to the post-dis fluent material (music) which would then be processed and represented differently to an equivalent stimulus in a passage of fluent speech. The thesis examines this proposition in the context of an explicit de finition of attention. The first half of the work investigates whether hesitations heighten two different aspects of listeners' attention: these are the immediate engagement of attention to post-dis fluent stimuli at the point they are encountered, and the continued attention to the representation of stimuli after they are encountered. In experiment 1, a speech `oddball' paradigm is used to show that event-related potentials (ERPs) associated with attention (MMN and P3) are affected by a preceding hesitation, indicating an immediate effect of hesitations on listeners overt attention. Experiments 2 and 3 use behavioural responses and eye-movements measures during a change-detection paradigm. These experiments show that there is also an effect on the listeners' attention to the post-dis fluent material after the initial presentation of the utterance. The second half of the thesis concerns itself with the timecourse of the attentional effects. It addresses questions such as: how long-lived is the attentional heightening and what is the attentional heightening trigger? Experiments 4{7 explore the relationship between the filler er and periods of silent pause that surround it. Behavioural (exp. 4{6) and ERP (exp. 7) results show that while extending the period of silence after the filler er does not affect the immediate engagement of attention, it will affect subsequent attention to the post-disfluent material: constituents that are not immediately preceded by the filler er are not attended to in an enhanced way. Together, these experiments confirm the proposition that hesitations heighten listeners' attention to upcoming speech. The thesis outlines the ways in which the components of this attentional heightening are differentially affected by interaction between the content and timing of the hesitations encountered. Attention has an important role to play in the processing of any stimulus. Using disfluency as a test case, this thesis illuminates its importance in language comprehension.

The phenomena of disfluency clusters have been examined in the speech of children who stutter (CWS) and children who do not stutter (CWNS). Little is known about disfluency clusters in the adult population. The purpose of this study was to examine the characteristics of disfluency clusters in adults who stutter (AWS). The participants were ten AWS ranging in age from 18 to 60 (mean age = 35), with a stuttering severity of 9 to 30% (mean = 19%). Each participant provided a conversational speech sample of at least 300 words. Analysis focused on disfluency type, utterance length, speaking rate, and perceptual measures. Findings indicated that utterances containing disfluency clusters were significantly longer than fluent utterances and the speaking rate of fluent utterances was found to be significantly faster than that of disfluent utterances. Collectively the results appear to support a linguistic interpretation of disfluency clusters. The clinical implications of the findings are discussed.

Natural spoken language is full of disfluency. Around 10% of utterances produced in everyday speech contain disfluencies such as repetitions, repairs, filled pauses and other hesitation phenomena. The production of disfluency has generally been attributed to underlying problems in the planning and formulation of upcoming speech. However, it remains an open question to what extent factors known to affect the selection and retrieval of words in isolation influence disfluency production during connected speech, and whether different types of disfluency are associated with difficulties at different stages of production. Previous attempts to answer these questions have largely relied on corpora of unconstrained, spontaneous speech; to date, there has been little direct experimental research that has attempted to manipulate factors that underlie natural disfluency production. This thesis takes a different approach to the study of disfluency production by constraining the likely content and complexity of speakers utterances while maintaining a context of naturalistic, spontaneous speech. This thesis presents evidence from five experiments based on the Network Task (Oomen & Postma, 2001), in addition to two related picture-naming studies. In the Network Task, participants described to a listener the route of a marker as it traverses a visually presented network of pictures connected by one or more paths. The disfluencies of interest in their descriptions were associated with the production of the picture name. The experiments varied the ease with which pictures in the networks could be named by manipulating factors known to affect lexical or pre-lexical processing: lexical access and retrieval were impacted by manipulations of picture-name agreement and the frequency of the dominant picture names, while visual and conceptual processing difficulty was manipulated by blurring pictures and through prior picture familiarisation. The results of these studies indicate that while general production difficulty does reliably increase the likelihood of disfluency, difficulties associated with particular aspects of lexical access and retrieval have dissociable effects on the likelihood of disfluency. Most notably, while the production of function word prolongations demonstrates a close relationship to lexical difficulties relating to the selection and retrieval of picture names, filled pauses tend to occur predominantly at the beginning of utterances, and appear to be primarily associated with message-level planning processes. Picture naming latencies correlated highly with the rates of observed hesitations, establishing that the likelihood of a disfluency could be attributed to the same lexical and pre-lexical processes that result in longer naming times. Moreover, acoustic analyses of a subset of observed disfluencies established that those disfluencies associated with more serious planning difficulties also tended to have longer durations, however they do not reliably relate to longer upcoming delays. Taken together, the results of these studies demonstrate that the elicitation of disfluency is open to explicit manipulation, and that mid-utterance disfluencies are related to difficulties during specific production processes. Moreover, the type of disfluency produced is not arbitrary, but may be related to both the type and location of the problem encountered at the point that speech is suspended. Through the further exploration of these relationships, it may be possible to use disfluency as an effective tool to study online language production processes.

Spoken language comprehension is impacted by the presence of disfluencies. It follows that there have been attempts to understand the underlying mechanisms that are responsible for these disfluency effects. Different accounts of disfluency processing have been proposed to explain these effects; the current thesis was directed towards exploring two standpoints of disfluency processing: the predictional and attentional accounts. Disfluency has been shown to modulate predictive processing, with a clear effect in the literature being that upon encountering disfluency listeners show a bias for unknown or discourse new referents (Arnold, Kam, & Tanenhaus, 2007; Arnold, Tanenhaus, Altmann, & Fagnano, 2004; Heller, Arnold, Klein, & Tanenhaus, 2014). The predictional account interprets this finding in terms of expectancy: according to this view, listeners expect speakers to produce harder-to-access words in situations where their linguistic performance is consistent with planning problems. Listeners are also more likely to remember words that follow a disfluency (Corley et al., 2007). The presence of disfluency has been shown to affect the attentional state of the listener, as indexed by attenuation of event related potentials to acoustically manipulated words post disfluency (Collard et al., 2008). These effects form the basis of the competing attentional account, which suggests that that upon encountering disfluency, listeners stop predicting about upcoming content and instead, employ heightened attentional resources to help them resolve the situation. In the first experiment, we aimed to distinguish between the predictional and attentional accounts by employing a visual world paradigm to investigate directly the underlying mechanism during comprehension. Participants were expected to show different fixation behaviour depending on which account was true. The main experiment provided some unexpected results, as the fixation behaviour seen would not have been predicted by either account. These results were further investigated in a number of post-hoc tests, testing participants sensitivity to the disfluency used in the main paradigm. The results observed were again inconclusive. Taken together these findings suggested that the mechanisms afforded by each account for disfluency processing may work in unison, with reliance on either attentional or predictive processing, or a mix of both, dictated by the demands of the task. In the remaining experiments (2-6) we focused on the attentional account of disfluency processing; we asked how disfluencies impact listener attention at a phonemic level. Pitt and Szostak (2012) demonstrated that the effect of phoneme manipulation is reduced when participants’ attention is explicitly directed to the ambiguous phoneme, with participants less likely to categorise an ambiguous item as a “word” under such conditions than otherwise. We applied this paradigm at the sentence level to investigate whether disfluencies induce heightened attentional focus at a phonemic level. Specifically, we compared the impact of a phoneme manipulation on lexicality judgements with; (i) attentional focus, and; (ii) disfluency presence. The initial experiments’ findings failed to replicate the attentional manipulation seen in the Pitt and Szostak study (2012) but results from the later studies suggested there is evidence that disfluency does drive listener attention but actually makes listeners more accommodating of the phoneme manipulation heard. These results are discussed in relation to the accounts of disfluency processing being tested.

Stuttering is a universal phenomenon that has been identified in ethnic and cultural groups around the world. While it has been suggested that attitudes toward stuttering are different for various cultural groups, knowledge of, and attitudes toward stuttering have not been studied extensively across cultures. The purpose of the present study was twofold: (1) to investigate the awareness and perception of stuttering for American children age 3 to 7 from diverse cultures, and (2) to compare those to findings of awareness and perception of stuttering for children from Israel and America. Sixteen children in four different age groups were asked to complete three different types of experimental tasks after watching a video of fluent and disfluent identical seal puppets. The participant’s awareness of disfluency was assessed through discrimination between fluent and disfluent speech and identification of the puppet who spoke like them. Perception was addressed through labeling and evaluation of fluent and disfluent speech. Results revealed that at as young as age 3 some children began to demonstrate accurate awareness of disfluent speech. However, the highest level of accuracy was not demonstrated in the majority of participants until age 7. In addition, results further revealed across all age groups that children were more accurate when discriminating between fluent and disfluent speech than identifying it. Similarities and differences between previous studies that have used the same experimental stimuli (i.e., Ambrose & Yairi, 1994; Ezrati-Vinacour et al., 2001) are discussed. The lack of diverse cultural participants and its resulting effects on the present study’s recruitment methodologies are also discussed. / text

Recently, research has postulated that stuttering is a motor disorder that results from brain abnormalities within the central nervous system. Based on evidence of numerous irregularities within various motor systems, it has been suggested that other motor domains may be comprised. In particular, research in individuals who stutter has found fine, gross, and visual-spatial motor impairment. These studies, though, are dated, have numerous methodological concerns, or yielded contradictory results. Thus, this study investigated whether motor skills in children who stutter (CWS) were compromised. Fine motor skills are important in a school environment because students are required to utilize these skills to complete various assignments and projects, such as cutting and folding paper. Gross motor skills are equally as important as children use these skills to move around their environment. Visual-spatial motor skills are vital for children as they are often required to copy notes off of the board. Deficits in any of these areas may have potentially harmful effects on school performance. Thus, in a school setting, school psychologists are a valuable asset, as they are trained to consult and work with "at risk" populations to prevent long-term problems. Given the potential motor deficits in CWS, school psychologists can intervene and provide appropriate accommodations to remediate any motor deficits. Participants included 12 CWS and 12 children who do not stutter (CWNS). Participants were recruited from a large urban school district and were administered the Bruininks-Oseretsky Test of Motor Proficiency-Second Edition (Bruininks & Bruininks, 2005; BOT-2). Parents completed a demographic questionnaire. One Way Analyses of Variance (ANOVAs) were calculated to compare group means. Results indicated that CWS performed poorer on all but one motor area. Given these results, when a child is identified with a disfluency problem, a broader consideration of issues that may be facing the child is warranted. In particular, school psychologists are in a position to intervene and provide appropriate services to an "at risk" population (i.e., CWS) by conducting a brief motor assessment to identify motor strengths and weaknesses. If warranted, school psychologists can provide accommodations and services to address any identified weaknesses in motor areas.

Spontaneous speech is replete with disfluencies: pauses, hesitations, restarts, and less than ideal deliveries of information. Disfluency is a topic of interdisciplinary research with insights from psycholinguistics, phonetics and speech technology. Researchers have tried to determine: When does disfluency occur?, Can disfluency be reliably predicted to occur?, and ultimately, Why does disfluency occur? The focus of my thesis will be to address the question of why disfluency occurs by reporting the results of analyses of disfluency frequency and the relationship between disfluency and eye gaze in a collaborative dialogue. Psycholinguistic studies of disfluency and collaborative dialogue differ on their answers to why disfluency occurs and its role in dialogue. One hypothesis, which I will refer to as Strategic Modelling, suggests that disfluencies are designed by the speaker. According to the alternative view, which I will call the Cognitive Burden View, disfluency is the result of an overburdened language production system. Throughout this thesis, I will contrast these two theories for an ultimate answer to why disfluency occurs. Each hypothesis attaches a functional role to a structural definition of disfluency and therefore in order to determine why disfluency occurs, I will contrast the structural and functional characteristics of disfluency. I will attempt to do this by analysing the dialogue behaviour in terms of speech goals and eye gaze behaviour a speaker is engaged in when they make certain types of disfluencies. A multi-modal Map Task paradigm was used in this thesis, in which speakers were asked to describe the route on a cartoon map to a distant confederate listener who provided either visual or verbal feedback. Speakers were eye-tracked during the dialogue and a record was kept of when the speaker attended to the listener’s visual feedback. Experiment 1 tested the visual feedback paradigm to establish its validity as a baseline condition. Speakers were found to make more disfluencies when they could interact with the visual feedback, suggesting disfluency is more common in interactive circumstances. Experiment 2 added verbal feedback to the experimental paradigm to test whether listeners react differently to the two modalities of feedback. Speakers made more disfluencies when the feedback was more complicated. Structural disfluency types were also observed to fulfil different functions. Finally, Experiment 3 manipulated the motivation of the speaker and found that Motivated speakers gazed more often and were more disfluent per opportunity than Control speakers suggesting that highly motivated subjects are more willing to engage in difficult tasks.

Made available in DSpace on 2016-04-27T18:11:46Z (GMT). No. of bitstreams: 1 nos limites da fluencia.pdf: 371976 bytes, checksum: 839d3536b0e16186b8c770eaf762bf4d (MD5) Previous issue date: 2005-08-23 / Objective: To investigate how cases of dysfluency, especially those within the limits between normal and pathological, have been diagnosed and the assessment criteria and therapeutic approaches used by speech therapists. Methods: Using a semi-guided interview, speech therapists talked about their conceptions and procedures with stuttering. The interview was recorded and literally transcribed to analyze data, which were organized in topics for discussion. Results: We could observe that stuttering had different approaches and, therefore, assessment and treatment procedures were equally different. Concerning the limits between normal and pathological dysfluent behavior, we noticed that not all interviewed therapists considered its existence. Conclusion: Considering that there are different types of clinical practice in speech therapy, subjects and complaints may be seen, assessed and treated differently. This clearly applies to speech fluency problems, given that there are professionals that focus on symptoms and not on the subject that presents them. Conversely, there are professionals that take into account the subject as a whole, without disregarding the context. Within such aspect, they value individual singularity, which does not comply with standardization of human behaviors into normal and deviant from normal. It was also concluded that the purpose of speech is to allow interpersonal communication, which takes place in different moments, and interruptions are accepted because speech fluency is not absolute in anyone / Objetivo: Pesquisar como os casos de disfluência, em especial aquelas que se encontram no limite entre o normal e o patológico, têm sido diagnosticado, quais os critérios de avaliação e qual a conduta terapêutica utilizada pelos fonoaudiólogos. Métodos: Através da entrevista semi-dirigida os fonoaudiólogos falaram sobre suas concepções e procedimentos sobre a gagueira. A entrevista, foi gravada e transcrita literalmente para análise dos dados, os quais foram organizados em temas para serem discutidos. Resultados: Pôde-se observar que a gagueira apresenta diferentes vertentes e portanto, procedimentos de avaliação e tratamento também distintos. Em relação a disfluência no limite entre o normal e o patológico, verificou-se que não são todos os entrevistados que consideram sua existência. Conclusões: Os tipos de clínica existentes na fonoaudiologia, permite que o sujeito e sua queixa seja visto, avaliado e tratado de maneiras diferentes. Nos casos da fluência de fala e seus problemas, isso pôde ser notado claramente, pois, há profissionais que focam o olhar no sintoma e não no sujeito que o apresenta. Por outro lado, existem profissionais que consideram o sujeito e sua totalidade, sem desconsiderar o seu contexto. Neste aspecto, valoriza-se a singularidade de cada um, o que não comporta padronizações dos comportamentos humanos, convencionando a normalidade e seus desvios. Conclui-se, ainda que o sentido da fala é permitir a comunicação com o outro, a qual se dá em momentos diversos, onde interrupções podem ocorrer, pois a fluência de fala não é absoluta em ninguém

Le chapitre 1 sert d’introduction à la thèse, pose les problématiques et les méthodes, remet en perspective les enjeux et annonce le plan suivi. Le chapitre 2 définit les principaux types de disfluences (cliniques et naturelles), résume les études principales conduites sur les disfluences, et présente les différents points de vue sur leur rôle dans le discours. Le chapitre 3 dresse l’état de la question sur le statut des deux pauses pleines (fillers) um et uh et montre comment plusieurs études récentes accréditent l’idée d’une différence pragmatique, voire fonctionnelle, entre ces deux "fillers", qu’il convient donc d’envisager comme des marqueurs. Le chapitre 4 revient sommairement sur le concept de "stance" (prise de position, évaluation), établit sa définition dans cette thèse et dans le corpus ATAROS, puis présente l’état de la question quant à la détection automatique de "stance" dans les corpus oraux. Le chapitre 5 caractérise les deux corpus étudiés, ATAROS et Switchboard (SWB), et établit leurs contributions. Ce chapitre présente les méthodologies d’annotation des corpus, les deux versions de SWB, ainsi que la méthode suivie pour construire une interopérabilité de ces deux corpus pour l’analyse de um et uh. Le chapitre 6 analyse la distribution et la durée des deux marqueurs dans SWB et ATAROS en fonction du genre des interlocuteurs, de l’authenticité de la conversation, et du nombre de conversations auxquelles les sujets participent. Ce chapitre montre que um et uh ont des durées et des distributions différentes et indique que les marqueurs ne sont pas utilisés au hasard. Le chapitre 7 se penche sur la production de um et uh dans SWB, et sur la perception des deux marqueurs en comparant les deux versions des transcriptions du corpus. Les principaux résultats montrent que um et uh sont plus souvent oublis que d’autres mots fréquents tels que les mots fonctionnels, et que les transcripteurs de SWB font plus d’erreurs sur uh que sur um, suggérant que um joue un rôle discursif plus important que uh. Le chapitre 8 interroge la relation entre la prise de position ("stance") d’une unité de parole et la présence et la position des marqueurs dans une phrase, et révèle que ces deux dimensions sont dépendantes. Le chapitre 9 évalue la relation entre la prise de position d’une unité de parole et la réalisation acoustique de la voyelle des marqueurs, comparé à la même voyelle dans d’autres mots monosyllabiques. Les résultats indiquent que les valeurs de "stance" affectent avec différents degrés la réalisation acoustique des marqueurs. Le chapitre 10 incorpore les résultats des expériences précédentes dans plusieurs taches de classification qui testent les traits les plus importants pour prédire automatiquement les valeurs de "stance" en fonction des paramètres correspondants à um et uh (traits lexicaux, positionnels et acoustiques). Ces expériences montrent que les traits pertinents aux marqueurs affectent la performance du système et que les meilleurs résultats de la classification sont obtenus lorsque les traits lexicaux um et uh sont présents, et lorsque leur position est prise en compte. Les résultats aussi indiquent que différentes propriétés acoustiques améliorent les scores de prédictions. Le chapitre 11 conclut la thèse en résumant les résultats des chapitres 6 à 10, en soulignant les impacts de cette recherche, et en indiquant les futures pistes de recherche. / Chapter 1 introduces the dissertation, establishes the research questions and the methodology, questions the stakes of studying the markers um and uh, and lays out the study organization. Chapter 2 defines the main types of disfluencies, clinical and naturally occurring, summarizes the state of the art on the topic, and presents the different positions on their discourse role. Chapter 3 establishes the challenges regarding the fillers um and uh and summarizes studies that support the idea of different pragmatic and functional roles, suggesting that they are markers rather than just fillers. Chapter 4 introduces the concept of ÔstanceÕ (i.e., evaluation, opinion), establishes the definition used in this study and in the ATAROS corpus, and briefly summarizes the state of the art on automatic stance recognition in spoken speech. Chapter 5 introduces the two corpora used in this dissertation, ATAROS and Switchboard (SWB), and establishes their contribution. This chapter presents the methodologies for the annotations, the two versions of SWB, as well as the methodology adopted to construct an interoperability between the corpora to analyze um and uh. Chapter 6 analyzes the distribution and the duration of the two markers in SWB and ATAROS depending on speaker and dyad gender, on the conversationÕs naturalness, and on speaker participation. This chapter shows that um and uh are different from each other, that they have different distribution and duration cues depending on the variables, and therefore indicates that they are not used randomly. Chapter 7 focuses on the production of um and uh in SWB, and on the perception of the two markers by comparing two transcription versions of the corpus. The results of this chapter show that um and uh are more often missed than other frequent words such as function words, and that SWB transcribers make more transcription errors on uh than on um, suggesting that um plays a more important role in discourse than uh. Chapter 8 investigates the relationship between stance and the presence and the position of um and uh in an utterance, and reveals that the presence and the position of the two markers is dependent with stance. Chapter 9 looks at the relationship between stance and the acoustic realization of the vowel of the markers, compared to the vowel of other monosyllabic words. The results indicate that the stance values affect the vowel realization to different extents. Chapter 10 consists of a classification experiment that incorporates the findings from previous experiments to find out which features pertinent to um and uh (lexical, position, and acoustics) improve the systemÕs performance. The experiments show that the features associated to the two markers impact the systemÕs performance and that the best results are obtained when the word unigrams um and uh are not filtered, and when their position is included. The results also indicate that different acoustic features improve the scores. Chapter 11 concludes the dissertation by summarizing the results from chapters 6 through 10, underlying the impact of this study, and addressing the future directions of this project.