=Paper=
{{Paper
|id=Vol-1419/section0016
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1419/section0016.pdf
|volume=Vol-1419
}}
==None==
https://ceur-ws.org/Vol-1419/section0016.pdf
Constraint Interaction in Language Processing and Acquisition
Chairperson
Sook Whan Cho
Sogang University, Seoul, Korea,
swcho@sogang.ac.kr
Speakers
Helena Hong Gao
Nanyang Technological University,
helenagao@ntu.edu.sg
Keum-Jeong Joo
University of Hawaii at Manoa,
noni1031@hanmail.net
(Co-authors: Kamil Ud Deen (U of Hawaii at Manoa) and Sook Whan Cho (Sogang University)
Say Young Kim
Nanyang Technological University,
sayyoung.kim@gmail.com
(Co-author: Helena Hong Gao (Nanyang Technological University)
Myung-Kwan Park
Dongguk University,
parkmk@dgu.edu
(Co-authors: Euiyon Cho, Wonil Chung, & Jong-Un Park at DonggukUniversity)
Jeong-Ah Shin
Dongguk University,
jashin@gmail.com
(Co-authors: Sun-Woong Kim (Kwang-Woon University), Gui-Sun Moon (Hansung University), Hye-Kyung Wee
(Dankook University), & Yoonjoo Na (Dongguk University)
Darcy Sperlich
National Kaohsiung University of Applied Sciences,
sperlich@kuas.edu.tw
Hongoak Yun
Konkuk University,
oaktreepark@gmail.com
(Co-author: Luca Onnis (Nanyang Technological University), Matthew Lou-Magnuson (Nanyang Technological
University), & Erik Thiessen (Carnegie Mellon University)
30
The objective of this symposium is to bring together Bilingual Children’s Flexibility in Paired
studies viewed as resulting from constraint interaction Conceptual Knowledge Retrieval
and cross-linguistic variation. In particular, a question Helena Hong Gao & Say Young Kim
is posed with respect to how different cues are weighted
in language processing and language acquisition when This study investigated 3- to 5-year-old Chinese-
they are in conflict. It is expected that competing English bilingual children’s conceptual knowledge and
approaches are presented and competition models in flexible interpretation of the words ‘big’ and ‘little’ in
psycholinguistic research are applied to individual English and ‘dà’ and ‘xiǎo’ in Mandarin Chinese.
languages and cross-linguistically while cross- Experiment 1 used Gao et al. (2014)’s design to require
fertilization between them may be sought for. Topics to children to interpret ‘big’ and ‘little’ both in reference
be addressed include the acquisition of adjectival to a medium-sized stimulus that was alternately
concepts, classifiers, and anaphora and the processing compared with a smaller stimulus and a larger stimulus
of verb phrase ellipsis, dependency relations, and (e.g., “Which one of these 2 balls is the big one?”). The
anaphora. Theoretical questions will be raised with a purpose of this experiment was to determine bilingual
focus on the existence and weighting of the constraints children’s conceptual knowledge base of BIG and
in pragmatic, semantic, and syntactic structures as well LITTLE in both languages. Experiment 2 examined the
as non-linguistic effects involving social cognition. It children’s conceptual knowledge retrieval with matched
is hoped that this symposium will provide an insight and unmatched visual input (e.g., A big image of a
into questions such as what competing motivations are small animal vs. a small image of a big animal). Even
at work, what factors determine the weightings of the the youngest children switched between interpretations
constraints, how different cues are in conflict, and how under normative conditions when lexical knowledge
different cues may be integrated toward a convergence. was applicable and ready to support. When flexibility
was required in the interpretation of big and little,
however, younger children perseverated on a single
Chinese Specific Lexical Development in Bilingual
interpretation under one condition (Exp. 1) and
Children: Linguistic and Social Factors
switched flexibly under another (Exp. 2). Results
Helena Hong Gao
support the statement of age-related conceptual
development with respect to adjective interpretation by
As a typological feature, noun classifiers in Chinese are monolingual children (Ebeling & Gelman, 1988, 1994;
unavoidable in everyday use of the language. For Gao et al., 2014) but also show a more flexible ability
example, “a book” in English must be expressed in in conceptual knowledge retrieval under the condition
Chinese as “yì běn shū”, which adds a classifier “běn” where metalinguistic and bilingual knowledge were
in between “a” and “book”. This compulsory structure both required.
requires Chinese speaking children acquire classifiers at
an early stage. Studies show that monolingual Chinese
References
speaking children can use common classifiers correctly
before schooling. Our question is: Can bilingual Ebeling, K. S., & Gelman, S. A. (1988).
Chinese speaking children use classifiers as effectively Coordination of size standards by young
as monolingual children? In this study, 30 English- children. Child Development, 59, 888–896.
Chinese bilingual children between age 7 and 12 were Ebeling, K. S., & Gelman, S. A. (1994). Children’s
recruited to participate in a noun-classifier association use of context in interpreting “big” and
task, in which children were required to count different “little”. Child Development, 65, 1178–1192.
types of commonly seen or used objects with a Gao, H. H., Zelazo, P. D., Sharpe, D., & Mashari, A.
classifier phrase (e.g., liǎng zhī bǐ “two pens”). Due to (2014). Beyond early linguistic competence:
the fact that counting the items unavoidably involved Development of children's ability to interpret
using classifiers, their application of the classifiers adjectives flexibly. Cognitive Development,
became natural output of their language use. An 32, 86-102.
interview of each child was followed after the task for Acquisition of the Korean anaphor caki: The
the purpose of understanding the reasons behind their interpretation of dative-marked subjects in
choice of classifiers. The results show that bilingual benefactive ditransitive constructions
Chinese speaking children used fewer classifiers and Keum-Jeong Joo, Kamil Deen, & Sook Whan Cho
they acquired classifier knowledge either via learning or
generating rules based on comparisons of object
properties or through imitation. Also, non-linguistic This study attempts to address the question of whether
factors, such as schooling, parents’ classifier Korean-speaking children and adult controls adopt (i) a
proficiency, father’s age, mother’s academic attainment, subject orientation or (ii) a nominative orientation in the
and income were found to influence a child’s classifier interpretation of the Korean reflexive caki ‘self’ which
proficiency. is known as a subject-oriented anaphor (Han &
Storoshenko, 2012; Montrul, 2010). However, the
subject orientation of the reflexive might be nominative
orientation. To test it, a Truth-Value Judgment Task
31
was conducted with 40 Korean-speaking children (age anaphors caki, caki-casin, and casin. In our study we
5;3-6;11) and 60 adult controls with two sentence types: used gender-denoting kinship nouns like siemeni
(1) a standard ditransitive sentence which has a 'mother-in-law' below to dictate either short- (SD) or
nominative and a dative, only the first of which is a long-distance (LD) binding of the anaphors at issue.
structural subject (e.g., Mickey-NOM Minnie-DAT self The results showed that in the LD relative to SD
ball-ACC kick-PST-SES) and (2) a benefactive condition, the critical gender-denoting noun after caki
ditransitive sentence with the benefactive marker –cwu elicited a positivity at the fronto-centro-parietal
which makes the sentence a biclausal one, with two region in the 600-650 ms interval (i.e., a P600 effect),
associated subjects—a nominative marked subject and a that after caki-casin also evoked a positivity at the
dative-marked subject (e.g., Mickey-NOM Minnie- centro-parietal region in the 550-650 ms interval, but
DAT self ball-ACC kick-BEN-PST-SES) (Hoshi, 1994; that after casin revealed a negativity at the anterior
Terada, 1990). The results show that both children and region in the 600-650 ms interval (i.e., an LAN effect).
adult controls strongly accepted a subject antecedent, We interpret these responses as indicating that in the
both in ditransitive and benefactive ditransitive LD relative to SD condition, the late (550-650 ms) ERP
sentences. Crucially, they accepted a dative antecedent responses associated with caki reflect complexity-
at a significantly higher rate in benefactive ditransitive related syntactic revision, those with caki-casin reflect
patterns than in ditransitive patterns. These findings syntactic repair, and those with casin reflect
contribute to a better understanding of young children’s morphological violation/ working memory difficulty.
awareness of (i) the subject orientation of caki, (ii) in
addition to a subject orientation, there is an additional
preference for nominative marked subjects over dative-
marked subjects and the biclausal structure of sentences
involving benefactive -cwu. In sum, these findings
support a syntactic analysis about caki as a subject
oriented anaphor.
References
Crain, S., & Thornton, R. (1998). Investigations I
Universal Grammar: A guide to experiments in
the acquisition of syntax and semantics.
Cambridge, MA: MIT Press.
Han, C.-H., & Storoshenko, D. R. (2012). Semantic
binding of long-distance anaphor caki in Korean. Syntactic structures in elliptical constructions:
Language, 88(4), 764–790. Evidence from structural priming
Hoshi, H. (1994). Passive, causative, and light verbs: A Jeong-Ah Shin, Gui-Sun Moon , Sun-Woong Kim
study on theta-role assignment (Unpublished Hye-Kyung Wee, Jong Un Park & Yoonjoo Na
doctoral dissertation). University of Connecticut,
Storrs.
A central issue in linguistic investigations of ellipsis is
Lee, C. M. (1973). Abstract syntax and Korean with whether ellipsis sites apply to syntactic representations. A
reference to English (Unpublished doctoral structural priming experiment in this study examines verb
dissertation). University of Indiana at Bloomington. phrase ellipsis in English, investigating whether
Montrul, S. (2010). Current issues in heritage language unpronounced ellipsis sites in verb phrases contain
acquisition. Annual Review of Applied Linguistics, syntactic structures, and they prime syntactic choice in
30, 3–23. subsequent sentence production. Structural priming refers
Terada, M. (1990). Incorporation and argument to people’s tendency to reuse the same structural pattern
structure in Japanese (Unpublished doctoral as one that was previously comprehended or produced
dissertation). University of Massachusetts, Amherst. (Bock 1986). In our experiment, we specifically use voice
mismatch constructions allowed in VP-ellipsis (e.g., The
dessert was praised by the customer after the critic did,
Repair, Revision, and Complexity in the Syntactic
too or The customer praised the dessert after the appetizer
Analysis of Dependency Relations for the Three
was, too) to see whether the ellipsis sites trigger access to
Korean Anaphors: An ERP-based Neuro-
syntactic structures and cause structural priming in
physiological Differentiation
subsequent sentence production. After encountering the
Myung-Kwan Park, Euiyon Cho, Wonil Chung, & previous sentences in the passive-active mismatch
Jong-Un Park condition, if active constructions are employed in the
subsequent utterances describing pictures (e.g., A
In an attempt to better understand how linguistic policeman chased a thief) in the picture-description task
dependencies are constructed in online sentence rather than passive constructions, we can argue that the
processing, we used event-related potentials (ERPs) syntactic structure of the elliptical constructions (i.e., the
to examine the processing of the three Korean critic did already) induces speakers to favor the
32
production of structures (i.e., A policeman chased a thief) order implicit knowledge. Crucial to the success of SL is
parallel to the syntactic representations in the ellipsis the role of individual experience with the world, and
sites, even though the syntactic structures in the ellipsis sensitivity to patterns of the perceptual environment.
sites and those in the antecedent clauses are mismatched Despite the importance played by experience, SL is
in voice. The results can be discussed within the syntactic tacitly conceptualized as a fixed skill, and few studies
approaches to ellipsis, not the nonstructural semantic have investigated how experience may shape statistical
approaches. learning. Our study is the first to adopt a pre-post test
design to modulate the effects of statistical learning on
References statistical learning itself. We provide both theoretical and
Bock, J. K. (1986). Syntactic persistence in language practical reasons for launching a research agenda on
production. Cognitive Psychology, 18(3), 355- statistical learning training, and evaluate our initial
387. results in the light of a parallel literature on cognitive
training in working memory.
The psycholinguistics behind reflexive pronoun
resolution in Vietnamese and Chinese
Darcy Sperlich
This presentation reports preliminary results of
reflexive pronoun acquisition of Vietnamese learners of
Chinese using a combination of interpretative judgment
and psycholinguistic data. The study investigates how
Vietnamese learners of Chinese interpret the Chinese
reflexive ziji (self) with the object to discover whether
syntactic or pragmatic processes are dominant.
Moreover, the Vietnamese participants have English as
a second language, hence how they understand the
English reflexive him/herself will be compared to their
Chinese processing. The underlying hypothesis is that
the Vietnamese will transfer their pragmatic reflexive
resolution strategies into their Chinese and English,
positive in the case of Chinese (as both languages are
pragmatic in nature), while vice versa for English.
Participants are given a self-paced reading test which
assesses timing spent per word, to time taken for a
judgment. An example sentence would be ‘John thinks
that David will give himself the apple’, and the
statement judged would be ‘The apple will be given to
John’ (another question for David would appear later).
The linguistic judgments are collaborated with
psycholinguistic measures of reaction timing,
confidence levels and knowledge source, along with
timing from the self-paced reading. Moreover, the
placement of the question is assessed (before versus
after the stimuli) to see how the question would affect
processing of the stimuli, as the reflexive is in a
sentence final position, after all potential antecedents.
In sum, it is hoped that this experiment will depict what
the main reflexive processing strategies are in the
languages considered, and if they are transferred.
Is statistical learning trainable? Preliminary results
Luca Onnis, Matthew Lou-Magnuson, Hongoak Yun &
Erik Thiessen
Statistical learning (SL) has been proposed as a core set
of cognitive mechanisms that support various higher-
order behaviors, from language to music and vision. It
involves mechanisms that extract information from the
environment, and abstract over stimuli to build higher
33