This article examines the Noticing Hypothesis – the claim that second language learners must consciously notice the grammatical form of their input in order to acquire grammar. I argue, first, that the foundations of the

hypothesis in cognitive psychology are weak; research in this area does not  support it, or even provide a clear interpretation for it. The problem of interpreting the hypothesis is much more acute in the area of language

acquisition. Partly because the hypothesis is not based on any coherent theory of language, it is very difficult to determine exactly what it means in this context,or to draw testable predictions from it.In the absence of specific predictions, research on form-focused instruction and feedback provideindirect tests, the results of which create additional problems for the hypothesis. The various problems can be eliminated or greatly reduced if the Noticing Hypothesis is reformulated as a claim that noticing is necessary for the acquisition of metalinguistic knowledge but not competence.

Recent years have seen a growing concern with the role of conscious processes in second language acquisition (SLA). This concern is frequently centered on the Noticing Hypothesis of Schmidt (1990; 1993a; 1994; 1995a; 1995b; Schmidt and Frota, 1986),which has been adopted by a large and probably growing number of researchers (e.g., Ellis, 1993; 1994b; Fotos, 1993; 1994; Fotos and Ellis, 1991; Harley, 1993; Larsen-Freeman and Long, 1991; Long,1991; Robinson, 1995; 1996; Zalewski, 1993). The hypothesis is a claim about how input becomes intake – that part of the input that is used for acquisition. It claims that conscious awareness (noticing)of grammar1 plays an important role in the process. In the strong form of the hypothesis, favoured by Schmidt (1990; 1993a; 1994;1995b), noticing is a necessary condition for learning. Other researchers might prefer a weaker version; that noticing is helpful but might not be necessary. I will assume the stronger version, though adoption of the weaker view would require only minor changes. The hypothesis also has strong and weak forms in another respect. In the weak version, learners need only be aware of the input in a global sense; they do not have to notice any details of its form. I will disregard this weak claim, for two reasons. First, in almost all discussion, advocates of the Noticing Hypothesis clearly favour the stronger view, that awareness of grammatical form is crucial. Second,the weak version is not interesting, because virtually no one would dispute it. Discussions of noticing are commonly directed against theories of unconscious acquisition, Krashen’s(1983; 1985; 1987) in particular. But no major theories, including Krashen’s, predict that learners benefit from ‘input’ that occurs while they are absorbed in some task that has nothing to do with the ‘input’. So I will assume that the Noticing Hypothesis requires conscious awareness of grammatical details rather than simply global awareness of input.

Noticing is often associated with the influential notion of consciousness raising (Rutherford, 1987; Sharwood Smith, 1981) or input enhancement (Sharwood Smith, 1991). But despite the (old)name, the idea does not imply a commitment to any particular view on the role of consciousness in learning. The latter is the concern of this article, so I will say little about consciousness raising.Proponents of noticing also give much attention to noticing the gap – learners’ awareness of a mismatch between the input and their current interlanguage (see especially Schmidt and Frota,1986). It is important to avoid confusion between this idea, which necessarily involves awareness, and the more general notion of a comparison between input and interlanguage. Theories of unconscious acquisition naturally hypothesize an unconscious comparison process. Schmidt and Frota (1986), in fact, presented noticing the gap as an adjustment of Krashen’s (1983) theory, the only difference being their additional claim that conscious awareness of the gap is a requirement. Thus, arguments that learners must compare input to their interlanguage grammar (e.g.,Ellis, 1994b) are not arguments for noticing.

I will argue that the Noticing Hypothesis is vulnerable in severalrespects. In the next section I challenge its foundations in cognitive psychology, concluding that it derives no support from research in this area. In the two following sections, I consider conceptual problems that arise in attempts to apply the concept of noticing to language acquisition. This is followed by a look at empirical problems. Finally, I suggest a reformulation of the hypothesis, the revised version being much narrower than current versions.Throughout the discussion I will focus on the work of Schmidt (1990; 1993a; 1994; 1995b), because he provides the cleareststatement and the most thorough defence of the Noticing.

阅读全文请访问：http://www.hss.nthu.edu.tw/~fl/faculty/John/Noticing%201998.pdf