Files in this item



application/pdfCHEN-DISSERTATION-2020.pdf (3MB)
(no description provided)PDF


Title:The acquisition of Mandarin by heritage speakers and second language learners
Author(s):Chen, Chung-Yu
Director of Research:Ionin, Tania
Doctoral Committee Chair(s):Ionin, Tania
Doctoral Committee Member(s):Montrul, Silvina; Shih, Chilin; Yoon, James Hye-Suk
Department / Program:Linguistics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):heritage language acquisition
second language acquisition
tone sandhi
aspect marking
relative clauses
Abstract:This dissertation investigates whether English-dominant heritage speakers (HSs) of Mandarin have selective advantages over proficiency-matched adult second language learners (L2ers) in several Mandarin linguistic phenomena, given the early age of acquisition (AoA) by HSs. Previous studies have found that HSs have an advantage over L2ers in phonology and core aspects of syntax, which develop before age three, but not in the domains of the lexicon, semantics, inflectional morphology, and syntax-discourse interface (see Montrul, 2012, 2016, for reviews). However, few studies have directly compared proficiency-matched HSs and L2ers across multiple linguistic domains, and none have done so for Mandarin. Thus, the goal of this study is to examine whether, when both Mandarin proficiency and the dominant language (English) are held constant, early AoA confers an advantage to HSs over L2ers of Mandarin, and whether this depends on the linguistic domain. Two broad research questions were asked: (1) Can HSs and L2ers of Mandarin whose dominant language is English fully acquire the properties of Mandarin that are different from or absent in English? (2) Do HSs have selective advantages over proficiency-matched L2ers, and does this vary by linguistic domain? To answer these questions, four Mandarin phenomena (tone 3 sandhi, aspect marking, relative clauses, and long-distance reflexives) were chosen because they are either absent or differently encoded in English, have different AoAs in monolingual Mandarin-speaking children, and are in different linguistic domains (phonology, morpho-semantics, syntax, and syntax-semantics interface). Three offline tasks were used for testing: a Tone Identification Task, an Acceptability Judgement Task, and a picture-based Truth Value Judgement Task. It is hypothesized that HSs will have acquired tone 3 sandhi, aspect marking, and relative clauses, as they are acquired by age five in monolingual children, but not long-distance reflexives, which are acquired after age eight when child HSs no longer receive extensive exposure to Mandarin. L2ers are hypothesized to have acquired Mandarin head-final relative clauses given that word order is relatively easy for L2ers, despite different headedness in English. However, L2ers may have difficulty with tone 3 sandhi because phonology is known to be difficult for them (e.g., Granena & Long, 2013), despite ample exposure to tones and tone 3 sandhi. Aspect marking may also be difficult for L2ers given that morphology presents a particularly challenging area (a ‘bottleneck’) for adult L2ers (Slabakova, 2008, 2014) and because there are cross-linguistic differences between English and Mandarin on aspect marking. Additionally, L2esr might find aspect marking difficult because predicates are not always marked with aspect markers, and there are restrictions on whether aspect markers can be combined with certain lexical predicates. Long-distance reflexives are expected to be difficult for L2ers due to a number of reasons, including English transfer, interface properties, presumably low frequency, and processing considerations. The results show that HSs were more native-like than L2ers in some domains, giving HSs a slight advantage. For tone 3 sandhi, HSs were slightly more native-like than L2ers in choosing more target-like T2T3 sequences for the T3T3 conditions. However, the advantage in tone 3 sandhi was not as robust as expected, likely due to undesired task effects that led even some Mandarin native speakers to not perform at ceiling in the Tone Identification Task. For aspect marking, HSs were more native-like than L2ers in that the latter seemed to be more subject to dominant language transfer from English progressive -ing, thus incorrectly allowing Mandarin progressive zai with achievement predicates. While relative clauses were expected to be easy for L2ers, HSs were slightly more native-like than L2ers in some but not all conditions (as in interpreting relative clauses and judging their headedness). Unlike Mandarin native speakers who accepted both long-distance and local readings of simplex reflexives ziji (though not at ceiling), neither HSs nor L2ers had acquired long-distance reading of ziji. Taken together, the acquisition of these different phenomena is influenced by multiple considerations, including transfer from English and domain vulnerability, with HSs having an advantage in those language phenomena that are acquired early in monolingual children. Frequency and processing considerations are briefly discussed.
Issue Date:2020-12-02
Rights Information:Copyright 2020 Chung-yu Chen
Date Available in IDEALS:2021-03-05
Date Deposited:2020-12

This item appears in the following Collection(s)

Item Statistics