[Abstract] Interactional approach is a very important theory in second language acquisition, including the input hypothesis and interaction hypothesis. This article first presents an exposition of these two hypotheses, and continues with a critical evaluation of them. Finally, the article concludes with the pedagogical implications of the approach.
　　[Key words] interactional approach hypothesis acquisition criticism
　　Introduction
　　In the past decade, the interactional approach has increasingly attracts people’s interest in second language teaching and learning. When teachers pay great attention to its evident strengths, they often overlook its weaknesses. Therefore, a critical evaluation is quite necessary for a better understanding of this approach.
　　This article is a discussion about the interactional approach in second language acquisition, focusing on Krashen’s input hypothesis and Long’s interaction hypothesis. In the first part, it presents an exposition of these two hypotheses through description and analysis of data. It explains the major claims of these two hypotheses and Long’s claims are further elaborated by the roles of negotiation of meaning in facilitating L2 acquisition. The second part of this article is a critical evaluation of these hypotheses. It first evaluates the evident strengths of this approach; meanwhile it points out its weaknesses by citing criticisms it has received. The major critique lies in the lack of definite empirical evidence to support these hypotheses. Finally, a conclusion is provided to state the pedagogical implications of this approach and the right attitudes we teachers should adopt towards it.
　　Ⅰ. Literature review
　　The Input hypothesis is first proposed by Stephen Krashen (1982, 1985). He claims that learners can acquire language by receiving input a little bit above their current level, that is, ‘i + 1’. According to him, a sufficient amount of comprehensible input is all that a L2 learner needs.
　　Krashen’s input hypothesis encouraged other researchers to examine more closely the linguistic environment for language acquisition, among them is Michael Long. Long (1981, 1983a, 1983b, quoted in Mitchell and Myles, p.167) also emphasizes the importance of comprehensible input, but he claims that input alone is not enough for acquisition to take place. One also needs interaction to facilitate language learning. Input is most effective when it is adjusted or modified through the negotiation of meaning. That is Long’s Interaction hypothesis.
　　Long argues that interaction can lead to acquisition in the following way:
　　Step1: Show that (a) linguistic/conversational adjustments promote (b) comprehension of input.
　　Step2: Show that (b) comprehensible input promotes (c) acquisition.
　　Step3: Deduce that (a) linguistic/conversational adjustments promote (c) acquisition.
　　(Long, 1985, p.378)
　　The first step is widely agreed that linguistic/conversational adjustments can promote comprehension. Linguistic adjustments are investigated mostly through the study of foreigner talk, the language that native speakers use when addressing non-native speakers. Foreigner talk is usually delivered at a slower pace; it is simplified and elaborated; and it is likely to use the ‘regular’ or ‘basic’ forms. All these strategies are believed to facilitate comprehension of input. The following example shows some of these features. The language is simplified by using shorter sentences, avoiding subordinate clauses and complex grammatical forms like question tags. Also a full form (‘will not forget’) instead of a contracted form (won’t forget) is used. Further, ‘on your way home’ is paraphrased as ‘when you are coming home’ to make the meaning clearer to the NNS (Ellis 1997, p.46).