Cross-Linguistic Transfer In The Acquisition Of L3 Word Order By Tibetan-Chinese-English Trilinguals

As Third Language Acquisition(TLA)gradually becomes an emerging research field independent of Second Language Acquisition(SLA),the research on trilinguals’ crosslinguistic influence(CLI)has also become the focus of most studies.In this thesis,we propose to investigate the cross-linguistic transfer patterns of Tibetan-Chinese-English trilinguals when acquiring English word order grammar by using both timed and untimed versions of Grammaticality Judgment Task(GJT).In the field of TLA transfer research,one of the central issues is that which language(s)would serve as the transfer source(no transfer source / exclusive L1 transfer / exclusive L2 transfer / combined L1 and L2 transfer)and what the transfer outcome(facilitative transfer / non-facilitative transfer)they would bring about.This ongoing doubt has been partially alleviated through the advancement of several L3 / Ln morphosyntactic transfer models or hypotheses,for instance,the Absolute L1 Transfer,the L2 Status Factor(L2SF),the Cumulative Enhancement Model(CEM),the Typological Primacy Model(TPM),the Linguistic Proximity Model(LPM),and the Scalpel Model(SM).Furthermore,research on this topic has identified a number of modulators that might determine the extent of L1 and L2 influence on the L3 / Ln acquisition process,such as overall(psycho)typological similarity among languages,structural proximity at the linguistic property level,L2 status as a foreign language,learners’ proficiency in their source language(SL)or target language(TL),task type,the dominant language of communication,etc.A number of TLA empirical investigations concerning the above issue has been conducted in order to test or expand the proposals of L3 / Ln transfer models or hypotheses,and examine the way in which modulating factors might influence the process of trilinguals’ language transfer.However,their results turned out to be inconsistent as they used varied language combinations,different target structures,and trilingual subjects with heterogeneous language backgrounds.Therefore,to contribute to our understanding of this issue,the present study sets out to examine the transfer source and outcome of Tibetan-Chinese-English trilinguals when they learn English word order knowledge,namely SVO order,attributive adjective placement,and numeral placement.Compared to previous studies,this study is innovative in that it uses a scarcely-tested typologically-distinct language combination,and takes task type into consideration as a possible deterministic factor of CLI.To this end,the following research questions are proposed:Research question 1: When Tibetan-Chinese-English trilinguals learn English word order knowledge,which language(s)(no transfer source / exclusive L1 / exclusive L2 /combined L1 and L2)serves as the transfer source(s)?Research question 2: Is task type a strong indicator of CLI when Tibetan-ChineseEnglish trilinguals learn English word order knowledge?To address these research questions,this study designed two experiments with both within-group and between-group designs.In experiment 1,30 Tibetan-higher-level(TH Group)and 30 Tibetan-lower-level(TL Group)trilingual participants were administered a timed grammaticality judgment task(TGJT,testing implicit knowledge)and an untimed grammaticality judgment task(UTGJT,testing explicit knowledge)presented in the Eprime 3.0 software.Similarly,30 Chinese-higher-level(CH Group)and 30 Chinese-lowerlevel(CL Group)trilingual participants were administered the same tasks in experiment2.The rationale of this experimental design is that by moderating the between-group proficiency level of L1 and L2,it is possible to examine the role of background languages on transfer patterns of L3 word order acquisition.Additionally,through administering different tasks for each group,the impact of different knowledge type could be tested on syntactic CLI.The experimental data were statistically analyzed by Generalized Linear Mixedeffect Model(GLMM)built in R software.Results of experiment 1 indicated that both L1 proficiency and task type had the main effect as TH Group scored significantly higher than TL Group on both implicit and explicit tasks,and both TH Group and TL Group obtained significantly higher scores on explicit task than on implicit task.However,in experiment 2,results showed that only task type had the main effect as both CH Group and CL Group obtained significantly higher scores on explicit task than on implicit task.On the other hand,L2 proficiency did not reveal a main effect as CH Group did not score significantly higher than CL Group on any type of the GJT tasks.Our empirical results indicated that only L2 Chinese was activated as the transfer source that facilitates L3 English word order learning.Therefore,our finding would uphold two TLA morphosyntactic transfer models,namely L2 SF and CEM.Meanwhile,L1 proficiency and task type might be important triggers that moderate the direction and degree of syntactic transfer.Possible explanations and implications are discussed in this thesis.