Search Abstracts | Symposia | Slide Sessions | Poster Sessions
When “Mary divorce John” becomes acceptable: An ERP investigation on the trend of transitivization in Mandarin
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Chia-Hsuan Liao1, Jia-Ying Tsai1; 1National Tsing Hua University
Introduction. Mandarin exhibits some flexibility in the number and the type of arguments that can occur with verbs. In social media, an increasing number of intransitive verbs are being used transitively (e.g., “Mary divorce John”), which linguists define as light verbs (Tsai, 2017; Lin, 2002). From a processing perspective, how these light verbs are parsed in sentence contexts remains unclear. To address the gap, we conducted an ERP experiment comparing the processing profiles of light verbs with those of unaccusative verbs (Perlmutter, 1978), a type of intransitive verbs that can never be used transitively (e.g., *“Fear disappeared participants”). If light verbs have indeed become transitive, no ERP differences would be observed between transitive and intransitive conditions. By contrast, if comprehenders still consider them intransitive, their ERP responses should be identical to the patterns observed with Unaccusative verbs: We predicted a P600 effect at the object noun phrase (NP) in the transitive conditions, as the presence of the object NP violated the subcategorization restrictions of the verb (Kielar, Meltzer-Asscher & Thompson, 2012). Design. The experiment used a 2X2 design, crossing Verb Type (Light, Unaccusative) with Transitivity (Transitive, Intransitive). Verbs were selected from social media and previous research (Tsai, 2017; Perlmutter, 1978). The lexical frequency of the verbs was matched between conditions. Each stimulus consisted of two clauses, always beginning with a NP1-Adverb-Verb-NP2 sequence. Transitive conditions were indicated by the presence of a comma after NP2, whereas Intransitive conditions had a comma after the verb, implying that NP2 should be interpreted as the subject of the subsequent clause. Participants (N=32) read 120 critical sentences (30 items per condition) and 120 fillers word-by-word, and performed a binary acceptability judgment task at the end of each sentence. ERP responses were time-locked to the onset of a verb and continued until the end of NP2. Results and Discussion. Both Light and Unaccusative in the transitive conditions elicited a larger LAN response at the verb, and a larger P600 response at NP2. The LAN effect indicates that the transitive use of these intransitives violated their subcategorization restrictions, whereas the P600 effect indicates that comprehenders preferred to interpret the post-verbal NP as the subject of the subsequent sentence. Since there were no qualitative differences between the two types of verbs, we suggest that comprehenders still consider light verbs intransitive. The goal of the study was to capture language change in progress, and we did observe significant variations at both the item and individual levels. Future work will focus on these variations by increasing the sample size and accounting for individual differences. By doing so, we can provide biological evidence for language change from a synchronic perspective. Additionally, further work may explore the impact of different linguistic contexts on the processing of these verbs, providing a deeper understanding of the mechanisms underlying language change.
Topic Areas: Syntax and Combinatorial Semantics,