You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

 
Poster E41, Friday, November 10, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Entrenchment in Chinese quadra-syllabic idiomatic expressions: A fMRI study

Shu-Kai Hsieh1, Tai-Li Chou2, Yu-Hsiang Tseng2, Chiung-Yu Chiang1, Chia-Lin Lee1,2, Te-Hsin Liu3, Chia-Rung Lu1, I- Ni Tsai3, I-Wen Su1;1Graduate Institute of Linguistics, National Taiwan University, 2Department of Psychology, National Taiwan University, 3Graduate Program of Teaching Chinese as a Second Language, National Taiwan University

Idiomatic expressions are commonly defined as fixed multi-word units whose semantic interpretation are mostly non-compositional, entrenched as functional pairings of form and meaning in speakers’ mind. This paper aims to explore a special type of idiomatic expressions with even length, called Quadra-syllabic Idiomatic Expressions (QIEs) in Chinese, and explain their variations with reference to the interaction of construction and lexical compositionality from an experimental perspective. QIEs are divided into two types: idioms (‘chengyu’) and prefabs. Idiom are culturally rooted and formed through ages of constant use, well compiled in dictionary and learned in school (e.g., hua4 xian3 wei2 yi2, ‘turn danger to safety’); while prefabs are more compositionally dependent, and can be understood as four-character lexical bundles (e.g., hao2 jiu3 bu2 jian4 “long time no see”). 96 Idioms and 96 prefabs are selected from a large corpus filtered by a certain threshold of frequency. Corresponding pseudo-idiom/prefabs with character replacements of different semantic distances are created (Near/Far). The total 384 words are rated by a different group of 40 native Chinese speakers on a 1–7 Likert scale regarding familiarity, comprehensibility, and the semantic distance between original words and replacements. 240 experimental stimuli are selected based on the comprehensibility rating results and divided into 4 groups: (1) Origin QIEs, (2) Near QIEs, (3) Far QIEs, (4) Random QIEs, with the top 30 Idioms and prefabs for (1) and (2); the bottom 30 Idioms and prefabs for (3). Participants are instructed to do a semantic congruency judgment during the presentation of a QIE. Two interesting results from fMRI are found. First, angular gyrus (AG, BA 39) is active in Origin QIEs and Near QIEs, but less obvious in Far QIEs and Random QIEs. This implies that origin QIEs and Near QIEs do operate semantic integration processing, but Far QIEs and Random QIEs do not. Second, left inferior frontal gyrus (IFG, BA45, 47) is significantly active in Near QIEs, but is relatively weak in Origin QIEs, Far QIEs and Random QIEs. IFG activation has been found in semantic tasks that demand higher retrieval processing. In our case, it is more active in processing Near QIEs, where we fill in the open slot of the original QIE scheme with a semantic related element. Namely, when processing Near QIEs, participants are working hard in inhibiting irrelevant properties or in choosing appropriate features. In sum, it is clear that the Origin and Near QIEs both operate AG, but IFG is only active in Near QIEs. Given the fact that the Origin and Near QIEs share the same construction, and the only difference between them is the manipulated open slot, we may assume that AG may reflect that schematicity and compositionality are both operated when we are processing meaning, and IFG indicates that there is an interaction between the construction-led schematic effect and compositional effect contributed by the composed semantic element in open slot. Hence, the results are in line with the recent cognitive constructionist proposal that the notions of schematicity and compositionality are not mutually exclusive.

Topic Area: Meaning: Lexical Semantics

Back to Poster Schedule