Understanding gesture-speech interaction through the lens of gestures' self-oriented functions: Gesture production and speech disfluency

People spontaneously gesture while speaking. The gesture-for-conceptualization hypothesis argues that gestures have self-oriented functions. Using representational gestures (e.g., leaving a small space between two hands to mean small) decreases cognitive load and helps individuals think and speak. Healthy individuals occasionally become disfluent when they speak. Do gestures facilitate cognitive processes and pave the way for more fluent speech, particularly when speaking is difficult? In this dissertation, I investigated multimodal language production in diverse groups, with a particular focus on gesture and speech disfluency. In three studies, I examined gesture-speech interaction as a function of age and bilingualism status as well as across bilinguals' first (L1) and second (L2) languages. In these groups, I investigated (1) speech disfluency and gesture production, (2) to which extent the two phenomena are related and temporally co-occur in spontaneous language, and (3) whether using gestures makes speech more fluent, complying with gestures' self-oriented functions. Chapter 2 examined whether younger and older adults revealed different types and rates of disfluencies and gestures in describing paintings and how age could affect the co-occurrence between gesture and speech disfluency, which has been rarely studied in the literature. Results indicated that younger adults produced higher proportions of filled pauses while older adults produced higher proportions of repetitions. Similarly, younger adults were more likely to produce representational gestures. Moreover, the proportion of speech disfluencies that occur with a gesture was significantly higher for younger than older adults. However, the two age groups were comparable in terms of the proportion of gestures that were accompanied by a speech disfluency. These findings suggest that gestures do not necessarily compensate for weak language skills. Instead, younger adults' language production system might be better at benefiting gesture modality to resolve temporary problems in speech. Chapter 3 investigated whether disfluency and gesture rates differed across 5- and 7-year-old Turkish monolinguals and Turkish-English bilinguals. I also examined whether children's gesture use was associated with their speech fluency in narrative production in both languages. Results showed that bilinguals were more disfluent than monolinguals, particularly for silent and filled pauses. Monolinguals produced more iconic gestures than bilinguals. Representational gesture frequency alone was not significantly associated with the disfluency rate in Turkish narratives. Instead, children's overall gesture frequency predicted disfluency rates only in Turkish, controlling for language proficiency and linguistic complexity. These findings together indicate that representational gestures' self-oriented functions might not be fully reflected in an emerging language system. Chapter 4 focused on bilingualism and multimodal language production in adults. I investigated disfluency and representational gesture rates of bilinguals in their two languages and whether and when representational gestures co-occurred with disfluencies across the two languages. Results demonstrated that bilingual adults were more disfluent in L2-English even after L2 language proficiency was controlled, suggesting that disfluencies might also stem from communicative intentions. Although bilinguals produced more representational gestures in their L2, this difference disappeared after L2 proficiency was controlled. The more participants were proficient in English, the more they used representational gestures in that language. Similarly, disfluency-gesture co-occurrences were more common in English. L2 proficiency was positively correlated with the likelihood of a disfluency being accompanied by a gesture. These findings suggest that individuals might use more representational gestures when they are disfluent if they have enough competency to detect their mistakes, which suggests a metacognitive background for gesture production. Overall, results from the three studies provide insight into typical multimodal language production through an investigation of disfluency and gesture patterns within and across populations. Despite their self-oriented functions, representational gestures do not compensate for weak language skills, but rather, they are closely linked to language proficiency. As a result, representational gestures' self-oriented functions might not be clearly reflected in children and older adults, which represent emerging and decaying language systems, respectively. In such populations, different gesture types might be orchestrated in the multimodal system, contributing to fluency. This dissertation theoretically informs the gesture-for-conceptualization hypothesis, suggesting a research avenue for representational gestures' self-oriented functions across age groups, bilingualism status, and one's L1 versus L2.