On the aquisition of grammar and meaning in instructed second language learning: a case study of the development of past verb forms by adult French learners of English as a foreign language.

A corpus of written English produced by three groups of adult French \nbeginning learners of English as a foreign language over a period of \napproximately eight months was examined for evidence relating to the \nacquisition of past tense forms and related meanings. \nThe findings provide evidence to support several hypotheses which can \nbe usefully grouped within a single framework which sees language \nacquisition as a process of hypothesis formation and testing whose \nconstraints are both first and second language in origin. These \nhypotheses can be summarised as follows: \n(1) Language learning involves the acquisition of a new system of \nexpressing meaning. As a result, the learner engages in a process of \nmatching linguistic form to underlying meaning both within and \nbetween languages. Only such a hypothesis, we believe, can satisfactorily \nexplain the apparently random variation that was observed in \nour subjects' acquisition of past tense. \n(2) Language transfer is thus necessarily a widespread phenomenon, \nconstraining learners' formation of hypotheses, but is itself constrained \nby the inter- and intra-language form-meaning transparency \nof the language item in question. In other words, whenever formmeaning \nrelationships are not wholly transparent, transfer is to be \nexpected. \n(3) Moreover, even when form-meaning relations are transparent, \ntransfer may take place due to the learner's shortage of processing \ncapacity. When this is lacking, learners tend to maintain communication \nby relying on existing procedural knowledge, which, at least in \nthe early stages, means well-established first language procedures. \nThis is because, even though humans can process at phenomenal speeds, \nthis is only possible with procedures which are solidly in place. For \nthe vast majority of language learners this implies that first \nlanguage procedures will always take precedence over weaker second \nlanguage procedures because they were later traced and are less \nfrequently used. \nUnder the circumstances, where demands are made on the learner, for \nexample, to produce language in real time, there will be a tendency \nto bypass second language networks and rely on first language \ncircuitry. Consequently, learners make use of their ability to \noperate a number of strategies (such as planning and monitoring their \nlanguage output) to produce comprehensible language. This ability, \nlabelled strategic competence, is identified as a key aspect in \nlanguage use in general.