Cognitive Mechanisms in the Perception of Sociolinguistic Variation
說明
201 p
附註
Source: Dissertation Abstracts International, Volume: 74-10(E), Section: A
Adviser: David P. Corina
Thesis (Ph.D.)--University of California, Davis, 2013
In our increasingly multicultural and multilingual world, an understanding of how we perceive language, dialects, and linguistic variation and the relationship these features have to language attitude, plays an increasingly important role in shaping social behavior and policy. This study, situated at the intersection of cognitive neuroscience, psycholinguistics, and sociolinguistics, uses multiple methodologies to investigate how individual speaker/listener differences influence the perception of linguistic variation and the social attitudes it engenders
The present study recorded participants' event-related brain potentials (ERPs) to investigate how and when information indexed by sociolinguistic variation is integrated during sentence processing. In a series of electrophysiological and behavioral experiments, participants listened to speech that varied in (ING) realization (working/workin'). In ERP experiment #1, Californian participants (n=28) heard short stories read by Californian speakers that were digitally manipulated to vary by speech register (formal/informal), variant (ING/IN'), and cloze probability (high/low). Findings showed (1) an N400 effect for cloze probability with low cloze words showing heightened negativities compared to high cloze words; (2) an N400-like effect for variant that was similar in latency, duration, and amplitude to the classic N400 cloze probability effect; with (3) increased N400 negativities for vernacular IN' variants compared to canonical ING variants; (4) increased N400 negativities to vernacular IN' variants preceded by informal register contexts compared to formal register contexts; and (5) greater N400 negativities for female compared to male listeners for sociolinguistic variant but not for cloze probability. According to these findings, word meaning and social information conveyed through spoken (ING) variants are integrated rapidly and concurrently into higher-order conceptual representations of meaning. However, when vernacular words variants are encountered during language comprehension, these processes are taxed. Results from ERP experiment #1 provide evidence that the cognitive mechanisms that support language comprehension are sensitive not just to what is said, but also to how it is said and who hears it
In experiment #2, Californian participants (n=19) listened to stories that were similarly manipulated to vary by register (formal/informal) and variant (ING / IN'). In order to further characterize the linguistic contexts under which sociolinguistic variants elicit N400-like negativities, passages were read by Californian and Southern speakers. We had reasoned that a stronger case for local linguistic contexts overriding broader sociolinguistic expectations could be made, if we could demonstrate an N400 like effect not just for vernacular IN' forms as in experiment #1, but also for canonical ING forms as well. Results showed (1) an N400-like negativity for IN' compared to ING words for Californian, but not Southern speakers; and (2) increased N400 negativities for formal ING passages uttered by Southern compared to Californian speakers. An additional analysis of the ERP data from experiment #2, investigated the role of listener gender. This analysis showed (1) N400-like negativities for females, but not males, listening to Informal IN' passages spoken by Californian speakers and Formal ING passages spoken by Southern speakers; and (2) heightened N400 negativities for females but not males listening to Californian speakers utter informal register contexts and for Southern speakers uttering formal register contexts
Taken together, the results of these experiments suggest that listeners have fine-grained representations of sociolinguistic and dialectal variation that are recruited during real-time language processing in order to help predict which variable word form a speaker is likely to utter. Importantly, these cognitive processes are further modulated by listener gender, with females showing particular sensitivity to sociolinguistic and dialectal norms. This work provides compelling neural evidence that the cognitive mechanisms that support language comprehension are dependent on what is said, how it is said, who says it, and who hears it
