guage instruction in the United States in the seventies and early eighties has been described as a cognitive approach. The theoretical that underlies the approach is that a language consists of a set of with an associated lexicon. It follows logically from the that foreign language students must learn rules of grammar. The suggested sequence is: study a (usually with instructor explanation), practice a (in grammar exercises), and then apply the in meaningful interactions in the target language. Krashen (13) has proposed a of second language acquisition in which the processing of input, rather than grammar instruction, plays the pivotal role. His hypothesis is that acquisition occurs when learners process in a low anxiety context. Learners presumably make use of a mental language acquisition device that allows them to store and produce utterances in the target language. Krashen does not attempt to specify how the acquisition process unfolds, but rather describes the conditions necessary for it to take place. He posits that the learner must be relaxed (have a low filter) and be focused on meaning rather than form. The must be comprehensible and in addition be at an + 1 level, that is, slightly more complex than the learner's current level of knowledge. An explicit knowledge of grammar by adults is said to be useful in only one way --as a monitor for self-correction under certain circumstances, to wit, that the learner know the rule to be applied, that the learner be focused on correctness, and that the learner have time to think about applying the to the output. Krashen has also suggested that grammar study may lower the affective filter for some adults and indirectly contribute to the acquisition process. The major implication of this input model is that learners' output is supposedly based directly on the they process and store. Children acquiring a first or second language seem to do what Krashen's predicts: there is usually a silent period in which children appear to build competence which is later displayed in their speech. While this appears to describe in broad terms how children acquire language, adult second language learners vary greatly in the strategies they use for second language acquisition. Some adult learners indeed do seem to rely heavily on for their output. They experience a rather long silent period and do not attempt much speech until they have built up some competence. As an example, consider DC, a Spanish-speaking learner of English living in San Diego for the past two years. DC has had a great deal of in English, but little of it is comprehensible (for example, he watches a great deal of television in English), and he has had even less at i (his level of comprehension) or i + 1. He has rarely attempted to speak English, preferring instead to rely on other more bilingual speakers of Spanish to translate for him. However, recently he has begun to speak some rudimentary English, and it is clear that his output consists of words and phrases he has heard at work (in a restaurant). Other adults appear to take a different approach to acquisition. As an example, consider RV, a native speaker of Dutch who has lived in the United States since he was fourteen. RV has managed to acquire a fair amount of competence in Spanish on a social basis solely The Modern Language Journal, 75, i (1991) 0026-7902/91/0001/052 $1.50/0 ?1991 The Modern Language Journal