An Inquiry Into Chinese Non-English Majors' Use Of Analogy In The Acquisition Of English Tense-aspect

Previous studies on L2 tense-aspect acquisition did not pay sufficient attention to the role of analogy or generalization in learners' attempts to infer the uses of L2 items for tense-aspect acquisition. Moreover, faulty analogy and overgeneralization, viz. the undesirable case of analogy and generalization, have been treated as separate concepts. As a consequence, they were too underpowered to explain how the L2 learner's mind works in the acquisition of L2 tense-aspect expression in particular and the acquisition of L2 grammar in general. The failure on the part of previous studies also suggests their lack of a robust underpinning of linguistic theories, because most of them drew on structural linguistics or Chomskyan linguistics.The present study, which was basically a qualitative study but adopted mixed methods, made both theoretical and empirical attempts to rectify these problems. Drawing on Langacker's Cognitive Grammar, this study attempted a cognitive linguistic account of English tense-aspect, which consisted of a sketch of the range of linguistic items pertinent to English tense-aspect expression, a tentative list of cognitive abilities vital to tense-aspect meaning construction, and a working theory of lexical and grammatical aspects, and which was later applied to the description of the English simple present and the English present progressive, i.e. the objects of acquisition the present study set out to investigate. This study also attempted a cognitive model based on Fauconnier and Turner's conceptual blending theory, which unified cognitive comparison, generalization, and analogy. In so doing, this study has highlighted two issues that are of paramount significance in L2 grammar acquisition: (a) how L2 learners identify similarities between L2 items that they are not familiar with and linguistic items that they have already acquired; (b) how L2 learners make inferences (or formulate hypotheses) about the use of L2 items, and thus the relation between L2 grammar acquisition and analogy/generalization was established.The empirical part of this study was conducted in the form of a revisionist grammaticality judgment test (GJT for short), which differed from traditional GJTs in terms of choice of the test items, and methods for data elicitation. The revisionist GJT had twenty-two non-English majors as participants, who were Chinese learners of English at intermediate level. The GJT which consisted of forty items that instantiated the uses of the English simple present and the English present progressive which are neither too general nor too idiosyncratic. The participants were required not only to make judgments of grammaticality and make corrections to the items judged to be ungrammatical (this was consistent with certain types of the traditional GJT), but also to state their reasons whenever they judged an item to be ungrammatical. The participants were also asked to translate into English two Chinese sentences which were designed to elicit their knowledge about lexical and grammatical aspects. Moreover, follow-up interviews were conducted with fourteen of the twenty-two participants in order to clarify certain respects of their GJT performances, and during their interviews they were required to answer three "global" questions regarding their acquisition of English tense-aspect expression. Hence, the empirical investigation yielded five types of data, namely, judgments of grammaticality, corrections made to the items judged to be ungrammatical, reasons given for conclusions of ungrammaticality, translation tasks, and interview results.The analyses of the data revealed certain propensities of the participants in learning English grammar, such as the propensity for tell-tale linguistic forms, the propensity for putting linguistic items of sub-regularity in comparatively more general categories, etc. And those propensities were interpreted in terms of the conceptual blending model so as to reveal the nature of analogy and generalization, and eventually to answer how learners identify similarities between linguistic units and make inferences about L2 use, and what elements might be misidentified, ignored or incorrectly inherited in the process of (over)generalization. It was concluded in the present study that elements most likely to be left out in L2 tense-aspect acquisition are semantic and discourse-pragmatic elements, i.e. elements that are not perceivable but crucial to identifying constraints on L2 use. This study also explained in terms of the conceptual blending theory how overgeneralization occurred in the participants' inferences about L2 use: the blended space, from which learners have constructed an L2 unit or a conceptualization in the way comprehensible to them, has inherited unwanted elements from the standard space, or has not inherited certain necessary elements, due to the faulty analogical alignment between the standard and the target (i.e. an ill-chosen standard) and/or the problematic projection of elements from the input spaces into the blended space. Hence the present study gave the time-honored and yet simplistic notion of overgeneralization a new and more appropriate interpretation, as well as a revelation of its cognitive nature.This study is a demonstration of the descriptive and explanatory power of cognitive linguistics in the study of L2 grammar acquisition and in the global enterprise of SLA research, and it proved the distinct superiority of cognitive linguistics to other linguistic paradigms in terms of scope, research perspective, descriptive instrument, and explanatory power.