Files in this item



application/pdfKIM-DISSERTATION-2019.pdf (2MB)Restricted Access
(no description provided)PDF


Title:Interpretation and processing of overt pronouns in Korean, English and L2-acquisition
Author(s):Kim, Eun Hee
Director of Research:Yoon, James Hye-Suk; Ionin, Tania
Doctoral Committee Chair(s):Yoon, James Hye-Suk
Doctoral Committee Member(s):Christianson, Kiel; Montrul, Silvina
Department / Program:Linguistics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Binding theory
L2 acquisition
L2 processing
Abstract:This dissertation investigates pronoun interpretation and processing of two languages (Korean and English), as well as in the second language (L2) acquisition of English by Korean speakers, with a focus on possible first language (L1) transfer. Using offline tasks and eye-tracking measures, the dissertation seeks answers to three primary research questions (RQs): 1) How are overt pronouns in Korean interpreted? 2) How are overt pronouns in Korean processed? 3) How do the properties of overt pronouns in Korean affect Korean-speaking English-learners’ interpretation and processing of English pronouns? Experiment 1 examined RQ 1 by testing the interpretation of Korean sentences including the overt 3rd person pronoun kunye ‘she/her’ in Korean, using two offline tasks. The focus was on whether the bound variable readings as well as the coreferential readings are available with kunye by testing different antecedent types – referential antecedents (names) and quantificational antecedents (every NP) – and whether and how kunye can take a clause-mate antecedent violating Principle B (Pr B) of Binding Theory (Chomsky, 1981). The results showed that Korean native speakers allowed kunye construed as the referential antecedent regardless of whether the antecedent was in the same local domain with the pronoun or not. However, Korean native speakers showed asymmetrical judgments with kunye construed as the quantificational antecedent: they rejected kunye when it took the local quantificational antecedent while they allowed it when it took the long-distance (LD) quantificational antecedent. These results suggest two things. One is that both bound variable and coreferential readings are available with kunye and the bound variable reading of kunye is constrained by Pr B. The other is that the instances of Pr B violation can also be found robustly due to the availability of coreferential readings and the suspension of the rule specifying the context where the coreferential readings can surface over the bound variable readings (Rule I, proposed in Grodzinsky & Reinhart, 1993). RQ 2 was addressed in Experiments 3 and 4 by examining whether and how Pr B is applied during the processing of the overt 3rd person pronoun in Korean, using eye-movement monitoring techniques while reading. More specifically, we tested if Korean speakers consider only the antecedents that are compatible with Pr B (i.e., LD antecedents) or they consider all the potential antecedents including the ones that are incompatible with Pr B (i.e., local antecedents) during the antecedent search process, by manipulating gender congruence between the pronoun and either of the two potential antecedents. We also tested if their processing pattern differs depending on the antecedent types (referential vs. quantificational). The results revealed the delayed gender incongruence effects with both antecedents, which indicates that both LD and local antecedents were activated during later stages of pronoun processing in Korean. The results also showed that there was no difference between the two antecedent types, which is unexpected. No ready explanation is present for why both antecedents are considered during parsing even with the bound variable reading of the overt pronoun (i.e., the pronoun with the quantificational antecedent), and hence future research should be conducted to follow up on this. Experiments 2, 5 and 6 addressed RQ 3 by testing how Korean-speaking learners of English interpret and process English pronouns and by comparing their performance with English native speakers’ performance and with their own performance in corresponding Korean tasks from Experiments 1, 3 and 4. In Experiment 2, the native speakers exhibited categorical judgments with pronouns as locally bound antecedents vs. LD bound antecedents, regardless of whether they were referential or quantificational, which suggests that their interpretation of the pronouns obeyed Pr B. However, the L2 learners’ responses were not consistent with Pr B: they overaccepted the English pronoun with the local referential antecedent, as they did with kunye in Experiment 1. Based on this similarity, it was suggested that their primary use of the coreferential reading in Korean may be at work and hence affect their interpretation of English pronouns. Experiment 5 also provided suggestive evidence that the pronoun processing pattern in Korean may influence their pronoun processing pattern in English when the antecedent was the referential NP. It was based on the similarity of L2 learners’ processing patterns in English as well in Korean such that the L2 learners activated both LD and local antecedents in the antecedent retrieval process. Their pattern was different from that of English native speakers who established the dependency between the pronoun and the LD antecedents immediately after encountering the pronoun and considered the local antecedents too at later stages of parsing. Interestingly, in Experiment 6, no such parallelism of L2 learners’ processing patterns in English and in Korean was found when the antecedents were quantificational. Instead, their processing pattern was parallel to that of English native speakers, who applied Pr B early in their processing and considered only the LD antecedents subsequently. The lack of suggestive evidence for the L1-transfer effect with the quantificational antecedents may be attributed to the different role of Pr B with referential vs. quantificational antecedents or the unresolved puzzle regarding the processing of Korean overt pronouns with quantificational antecedents. Taken together, these results indicate that coreferential reading of the overt pronoun in Korean is readily available due to the suspension of Rule I and it influences not only how the Korean overt pronoun is processed real-time but also how Korean-speaking learners of English interpret and process English pronouns. Implications are discussed in relation to the role of the L1 in L2 acquisition/processing, as well as the status of the overt 3rd person pronoun in Korean and the general sentence processing model available to a head-final language.
Issue Date:2019-06-26
Rights Information:Copyright 2019 Eun Hee Kim
Date Available in IDEALS:2020-03-02
Date Deposited:2019-12

This item appears in the following Collection(s)

Item Statistics