UC Berkeley

How do neighboring languages come to resemble one another, regardless of their genetic affinity? The literature is rife with examples of areas all over the world where languages have converged and share multiple features, but what are the exact mechanisms by which these large-scale effects come about? This dissertation integrates viewpoints from phonetics, phonology, historical linguistics, sociolinguistics, language contact, and multilingualism to investigate this question. Ultimately, the effects that we see from language contact are a form of language change involving influence from other languages. However, as languages cannot directly influence one another, the question cannot be answered without analyzing what individuals are actually doing in language contact situations. Naturally, these situations entail that they are speaking and listening to multiple languages, and so it is the production and perception of language that this dissertation will probe.

Previous work on language contact has acknowledged bilingualism as a prerequisite for contact-related changes that permeate beyond the lexicon and derives language contact effects as deriving from processes of imposition of one language's grammar onto another within bilinguals; however, much of this work tends to be restricted to theorizing these processes from the surface outcomes of language contact. On the other end, there is a wealth of experimental literature exploring how bilinguals use cues in their first and second languages; however, much of this work is focused on the question of what bilinguals do, without expanding to the broader context of how the results may have implications for understanding contact effects on language change. As such, this dissertation seeks to unify these two bodies of literature.

The sound change of interest is tonogenesis, the emergence of contrastive tone, which is well-known to arise from segmental contrasts through a combination of articulatory and perceptual factors involving co-occurring phonetic cues in phonological contrasts. The existence of co-occurring cues in a phonological contrast is of particular interest from the language contact/multilingualism angle, as the informativity of a cue may be influenced by the informativity of that cue in another language spoken by the multilingual. As a case study, I examine the realization of phonological register in a quadrilingual Kuy community in Northeast Thailand as a case study. Members of this community speak two non-tonal languages, Kuy and Khmer, and two tonal ones, Lao and Thai. As pitch is a cue common to both register and tonal contrasts, I explore how Kuy speakers' usage of pitch aligns with their usage of these languages by carrying out a production and perception study and analyzing the results in the context of sociolinguistic data. I hypothesize that greater usage of tonal languages will correlate to greater usage of f0 (fundamental frequency) in the Kuy register contrast. I also discuss how these language patterns arise against the background of increasing pressures to use Standard Thai, due to a combination of changing schooling patterns, greater mobility between provinces, and an overall higher degree of centralization.

Chapter 1 provides the relevant sociolinguistic background and language contact situation of the quadrilingual situation in the Kuy community and review literature related to sound change, language contact, and multilingualism.

Chapter 2 explains the context in which I carried out linguistic fieldwork and the methods I used to process and analyze the sociolinguistic and phonetic data from the studies.

Chapter 3 describes the production experiment I carried out. Participants embedded target Kuy words differing minimally in register in a carrier sentence. The results show stark gender differences, with female speakers showing a positive correlation between usage of tonal languages and usage of f0 in production of the Kuy register contrast, and male speakers lacking any correlation. The relationship between usage of tonal languages and usage of voice quality was more complicated, including a mix of positive, negative, and null correlations.

Chapter 4 describes the perception experiment I carried out. Participants listened to perceptual stimuli of Kuy minimal pairs manipulated for f0 and voice quality and identified the register of each stimulus in a forced-choice task. The results show a positive correlation between usage of tonal languages and usage of f0 in perception of the Kuy register contrast for both female and male listeners. There was no relationship found between usage of tonal languages and usage of voice quality.

Chapter 5 explores the relationship between the production and perception results by examining the results in tandem in participants who partook in both studies. The results show a positive correlation between usage of tonal languages and usage of f0 overall and no significant relationship with usage of voice quality. This chapter discusses potential explanations for asymmetries in the findings with regards to gender and for differences in the production and perception results. I also analyze and discuss the usage of f0 in the register contrast in the context of three other ongoing sound changes that may also be linked to contact with Thai and Lao.

Finally, in Chapter 6, I summarize the findings and expand on the broader implications and propose that language contact effects can potentially be understood as the enhancement of preexisting features in a language due to the informativity of those features in another language. While this dissertation focuses on understanding how sound change can be shaped by language contact, changes at any level of language can be understood under the same framework. The results of these studies suggest that large-scale changes in social situations shift individual language patterns and trigger micro-level shifts in cue usage in a linguistic contrast. These shifts can bias languages to change in certain directions and set the stage for the macro-level language contact effects that we see over time.