左右半腦差異與句法結構複雜度之學習——以事件相關腦電位分析人工語法處理狀況

Right Hemispheric Syntactic Processing as a Function of Saliency of Syntactic Regularities: an Event-Related Potential Study of Artificial Grammar Learning

顏瑄慧(Shiuan-Huey Yen)

指導教授 ： 李佳霖

國立臺灣大學/文學院/語言學研究所/碩士(2019年)

https://doi.org/10.6342/NTU201900033

摘要

語言的特性是區別人類與動物的重要指標，其中，句法處理更為高階困難。大腦在分析句法結構時，在大多數的人傾向於側化至語言中樞的左腦處理，在事件相關腦電位（Event-Related Potentials，ERPs）的研究中，發現右腦也有相似於左腦，在遇到不合文法的結構時，產生分析句法結構的腦電波反應（P600）——不過截至目前的研究中，左右兩腦都顯現句法結構分析的腦電波，大多是在語言能力以及語言學習成效較差的人身上觀察到。右腦之於句法處理，是作為左腦平行產生，兩腦彼此競爭、干擾，造成學習成效不佳？抑或在遇到複雜句法結構、學習有所困難時，作為輔助與支援？目前文獻對於右腦在句法處理上擔任的角色尚未釐清。為了觀察右腦參與句法處理的功用、以及左右腦在遇到句法結構或學習結構困難上的互動，本研究採用事件相關腦電位（Event-Related Potentials，ERPs）作為量測工具，紀錄無左利家族史的年輕成人受試者在人工語法學習（Artificial Grammar Learning）上左右腦處理的反應。研究由兩個實驗組成，實驗分為學習階段與測驗階段。於學習階段，受試者們在其中以單耳交替聆聽（alternate monaural listening）依照非相鄰依存（nonadjacent dependencies）文法結構設計、三個字串組成的句子；於測驗階段，實驗更動其中一半句子的非相鄰依存文法結構配對為不合規率之配對，並請受試者進行文法判斷測驗。實驗一結果以命中率與假警報率差（命中率（hit rate）減去假警報率（false-alarm rate），以H-F稱）區隔學習成效佳的受試者與學習成效較不佳的受試者。事件相關腦電位結果顯示，學習成效佳者（H-F = 0.91，共17人）呈現左腦側化句法處理反應，僅於右耳呈現刺激材料時顯示P600的文法效果。而學習成效較不佳者（H-F = 0.06，共15人）則左右半腦雙側化的句法處理，在左右兩耳播放刺激時皆有穩定的P600文法效果。不同於實驗一，實驗二將非相鄰依存文法結構的呈現分成四個階段，並僅分析最後學習成功（最後兩階段再認分數 > 0.8）的受試者。同時，實驗二加入不同的文法結構難度，分為非相鄰依存文法配對明顯的簡易組（high variability group）、與非相鄰依存文法配對不明顯的困難組（low variability group），以期觀察大腦相應文法結構複雜度的不同表現。於實驗二，簡易組（H-F = 0.96，共16人）於事件相關腦電位結果呈現左腦側化之P600句法處理反應；困難組（H-F = 0.95，共16人）則於左右腦皆呈現句法處理的反應。綜合實驗一與實驗二的結果，本研究推論在句法結構分析的過程中，在句法結構較簡單、或學習成效佳的情況，句法處理傾向於側化至左腦，然在遇到較困難的結構時，右腦則以互補的形式呈現P600，輔助句法結構學習。

關鍵字

語言習得 ； 句法處理 ； 人工語法學習 ； 統計學習 ； 左右半腦差異 ； 個別差異 ； 事件相關電位 ； P600

並列摘要

Syntactic processing, as one special characteristic of human language, is strongly lateralized to the left hemisphere (LH). In Event-Related Potentials (ERPs) studies, the right hemisphere (RH) has been found to elicit P600 effect to grammatical errors, showing equivalent syntactic processing ability to the LH. However, the RH P600 effect are often observed in individuals with lower sensitivity to grammatical structure. Under more challenging syntactic processing conditions, the RH might either compete with or attempt to assist the LH. In order to have a better understanding of the interaction between the LH and RH in processing of more complex syntactic regularities, we conducted 2 Event-Related Potentials (ERPs) experiments of young right-handers’ artificial grammar learning. The participants monaurally learned and made grammaticality judgment of a set of 3-item strings generated according to pre-determined artificial grammar rules. In experiment 1, LH-only P600 grammaticality responses are found in successful learners (Hit-False Alarm differences (H-F = 0.91, N = 17), while bilateral P600 effects are found in less successful learners (H-F = 0.06, N = 15). Similar to experiment 1, experiment 2 presented the materials in experiment 1 in 4 sessions and focused only on the successful learners (whose H-F > 0.8 in the last 2 sessions) to have a better look at the brain responses. Experiment 2 further manipulated the conditions as salient (i.e., easier to learn) and less salient (i.e., difficult to learn) regularities. In the salient group (H-F = 0.96, N = 16), LH-only P600 grammaticality responses are observed, yet bilateral P600 effects are observed in the less salient group (H-F = 0.95, N = 16). Results of the two experiments implicate that RH P600 effects of syntactic processing can be compensatory in more challenging learning tasks.

並列關鍵字

language acquisition ； syntactic processing ； artificial grammar learning ； statistical learning ； hemispheric differences ； individual differences ； ERPs ； P600 ； monaural presentation

參考文獻

1. Bakker, D. J., Smink, T., and Rietsman, P. (1973). Ear dominance and reading ability. Cortex, 9, 301-312.

2. Bakker, D. J. (1970). Ear-asymmetry with monaural stimulation: Relations to lateral dominance and lateral awareness. Neuropsychologia, 8, 103-117.

3. Banich, M. T. (2003). The divided visual field technique in laterality and interhemispheric integration, ed. K. Hugdahl, Experimental Methods in Neuropsychology, 47-64. Norwell, MA: Kluwer Academic Publishers.

4. Barwood, C. H. S., Murdoch, B. E., Whelan, B. –M., Lloyd, D., Riek, S., Sullivan, J. D. O’, Coulthard, A., and Wong, A. (2011). Improved language performance subsequent to low-frequency rTMS in patients with chronic non-fluent aphasia post-stroke. European Journal of Neurology, 18, 935-43.

5. Berlucci, G. (1983). Two hemispheres but one brain. Behavioral and Brain Sciences, 6, 171-173.