This article reviews research on working memory (WM) and its use in second language (L2) acquisition research. Recent developments in the model and issues surrounding the operationalization of the construct itself are presented, followed by a discussion of various methods of measuring WM. These methods include word and digit span tasks, reading, listening and speaking span tasks. We next outline the role proposed for WM in explaining individual differences in L2 learning processes and outcomes, including sentence processing, reading, speaking, lexical development and general proficiency. Key findings are that WM is not a unitary construct and that its role varies depending on the age of the L2 learners, the task and the linguistic domain. Some tests of WM may in fact be tests of differences in ability to attend to aspects of the L2. Future research will focus on matching tests of WM more closely with linguistic tasks and using more standardized, replicable measures of WM in new areas including writing in non-alphabetic scripts, instructional interventions and cognitive neuropsychology.
This study investigated how adult learners of English as a second language (ESL) process sentences containing verbs that are temporarily ambiguous in interpretation between a main verb and a reduced relative clause. Seventeen Chinese, 17 Korean or Japanese, and 17 Romance learners with advanced ESL proficiency and a comparison group of 17 monolingual native speakers (NSs) of English provided word-by-word reading times for 6 sentence types. The evidence showed that they used both verb subcategorization information and post-ambiguity cues to resolve main verb/reduced relative clause ambiguity. The data also indicated that bad post-ambiguity cues misled some ESL users more than others, differences that can be attributed to their first languages (L1s). These results suggest that (a) like native speakers, ESL speakers are sensitive to the complex interaction of information sources when parsing a sentence; (b) adult ESL learners are influenced by typological properties of their L1s that are linked to L1 parsing strategies when processing ESL.
Current theories of morphosyntax propose that features play a key role in determining some aspects of word order and in the argument structure of verbs. Thus, knowledge of language consists of features and mechanisms that ensure that elements in a clause are assembled correctly for interpretation and phonological processing. Therefore, the nature of the grammar of an L2 can be explored by assessing whether features and the mechanisms through which they are assembled and checked is 'impaired' or not. This article assumes that processing performance can provide a window onto this grammatical competence. Evidence is presented from online processing of English by thirty Chinese-speaking, twentyeight Japanese-speaking, and forty-six Spanish-speaking participants which shows that the basic mechanisms of grammar remain intact for L2 learners in spite of inferior performance on judgements of those same sentences. Data suggest that working memory, as measured by the reading-span test (Daneman and Carpenter 1980), is not a source of individual variation in online L2 performance, whereas word span might be.
This article expands on Juffs & Harrington's (1995) investigation of the parsing performance on wh‐movement sentences by Chinese‐speaking learners of English. We compare the difficulty L2 learners have in parsing subject wh‐traces in embedded finite and nonfinite clauses with the problems they have in parsing Garden Path (GP) sentences. Using the moving window technique (Just, Carpenter, & Woolley, 1982), 25 Chinese‐speaking learners of English supplied word‐by‐word reading times and grammaticality judgments on a range of wh‐extraction structures and GP sentences in English. Analysis of the error and accurate judgment data and the word‐by‐word reading profiles supports the hypothesis that L2 learners of English may have a parsing, rather than a competence, deficit in judging grammatical wh‐extraction.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.