A Study Of The Effects Of Working Memory On Chinese L2Learners’Short-term Sentence Recall And Sentence Comprehension

The relation of working memory and sentence processing is one of the focuses in the psychological linguistics field. Much of the research has been done on sentence processing in L1; yet little is known about L2learners’ processing performance. Besides this, researches in L2mainly focus on individual difference shown on the influence of working memory on the comprehension of complex sentences, such as the comprehension of complex sentences. It is still not clear about how other factors are constrained by memory capacity in L2sentence processing. Since sentences are processed at several different levels and along different information aspects (such as phonological, lexical, syntactic and conceptual information), it is necessary to take into account the various factors related to sentence processing to further examine L2learners’cognitive mechanism.On the basis of the previous research results, the present study attempts to explore the role of working memory in L2sentence processing, containing the effect of working memory (WM) on phonological and conceptual factors as well as its influence on syntactic factor. The experimental study consists of three tests. The first test is WM span test in which the subjects were required to recall the final word of each sentence and meanwhile judge the meaningfulness of each sentence. They were divided into high and low span groups according to their working memory span. After that, they were asked to take short-term sentence recall task under both visual and auditory presentation, in which Potter’s lure-intrusion paradigm was employed. The explanations behind this experiment mainly concentrate on two kinds of assumptions. Potter (1990) claimed that short-term sentence recall is mainly based on conceptual representations, while Rummer (2001) believed that Potter and Lombardi went a step too far with their conceptual regeneration hypothesis and their experiments were not aimed at demonstrating that phonological information is also involved in sentence recall, at least with auditorily presented sentences due to the involvement of surface memory. Thus, in order to explore the role of working memory in sentence processing, especially the effect of the phonological loop, the subsystem of working memory, on phonological encoding and short-term store and its influence on conceptual regeneration, both phonological information and conceptual representations should not be neglected because of their contributions to short-term sentence recall. Finally, an on-line sentence comprehension test was carried out to examine the effect of working memory on syntactic factor in L2sentence processing.The results can be concluded as following:firstly, a recall advantage for high span group under both auditory and visual presentation can be observed, suggesting that working memory span does affect short-term sentence recall task, and meanwhile the author also found that, in general, L2learners are not naturally supported by phonological information to the same extent as native speakers in sentence processing due to phonological interference, which may on the other side suggest the importance of phonological recoding in sentence processing. Secondly, more lure intrusions appeared for low span subjects than those for high span subjects, which may be attributed to the episodic-specific links from short-term information to long-term memory concepts. This result may to some extent suggest that potter’s conceptual STM is more compatible with Baddeley’s episodic buffer (2000) in terms of function.Thirdly, as to the role of working memory in syntactic processing, the result supports previous studies on syntactic processing:the more difficult the syntax, the more resources the memory demands. In addition, the author found that high span subjects, unlike low span subjects, are less influenced by syntactic operations, implying that high span subjects process sentences in a parallel manner. Finally, as expected, correlation test has shown that working memory span has a positive correlation with both short-term recall and sentence comprehension tasks, while at the same time, we find that sentence comprehension task is correlated with recall of auditorily presented sentences, while it has no positive correlation with recall task under visual presentation. The author attributes this to the function of phonological loop in the process of comprehension. But the reason hasn’t been deeply analyzed and left for future study.