4 Language acquisition: methods and metaphors 4.1 Introduction The use of linguistic frameworks in the modeling of first language acquisition has a long pedigree, and has influenced kinds of frameworks applied to the study of second language acquisition (see Stern 1970:57ff; Brown1980:58-61). Many of the issues prominent in the study of one kind of acquisition continue to overlap with those in the other. But whatever the initial influence first language acquisition studies may have had, or continue to have, on second language studies, there is no doubt that second language acquisition (SLA) has developed in its own right as a branch of applied linguistics. In many respects, SLA may claimed to be a classic example of applied linguistics, in that its proponents have fought hard battles to establish its dependence both from a one-way dependence on linguistics (Widdowsons linguistics applied; see section 1.2), and from being seen merely as a branch of educational psychology. Some of its most faithful adherents see it as a rationalist, scientific enterprise, with its own theoretical base and its own avenues of enquiry, in which methods associated with the natural sciences play a major role, for example, laboratory-condition experiments (see the special issue of Studies in Second Language Acquisition, 1997,19(2), and section 4.2 ). But also, of late, of late, debates have raged the status of SLA investigations. By no means everyone agrees that SLA should model itself on the hard sciences, and many see SLAs future as bound up with more humanistic and holistic approaches. Some question whether SLA has contributed anything at all in a real, practical sense to second language pedagogy. In this chapter we shall examine the broad evolution of SLA as a sub-field within applied linguistics and try to relate the debates that fuel its progress to th fundamental issues raised in the previous chapters. This introductory chapter cannot do full credit to what is a very large, complex undertaking; a fully comprehensive survey of the field of SLA can be found in R.Ellis (1994). 第四章 语言习得：方法与隐喻 4.1 引言 在建立第一语言习得的系统化模型中，语言学框架很早就被使用了，它还影响了应用于第 二语言习得的语言框架。 （参见 Stern 1970:57ff; Brown1980:58-61） 。许多在一种语言习得研 究中重要的议题，在其他习得研究中也很重要。但无论第一语言习得研究会产或继续产生 什么影响，毫无疑问，在第二语言研究方面，第二语言习得已经凭自己能力成为应用语言 学的一个分支，在许多方面，第二语言习得可能都是应用语言学的一个典型实例，因此， 它的支持者们经过努力斗争，确立了第二语言习得的独立地位，而不是单方面依靠语言学 （Widdowson 的“语言学的应用” ，参见第 1.2 节）或者仅仅被视作教育心理学的分支，一 些第二语言习得理论的最忠实追随者认为，这种理论是理性主义的科学的工程，并且有自 己的理论基础和探索手段。理论以这种方式与自然科学比如实验条件下的经历（参见第 二语言习得研究的特别议题，1997,19(2),和 4.2 节）相联系，占有主要地位。但是，最近 争论已经涉及到第二语言习得调查的地位了。人们绝不会相信第二语言习得会成为硬科学， 许多人认为第二语言习得的未来是与更人性化的，更全面的途径相联系的。一些人怀疑第 二语言习得是否一点也没从真实、实际的角度给第二语言教学法作贡献。在本章中，我们 将检查第二语言习得作为应用语言学的一个下级领域的广泛的创新性，并试着把促进它进 步的争论和在上一章中引起的基本问题联系起来。R.Ellis（1994）发现的第二语言习得领 域的一个充分可理解的调查指出这篇引入的章节不能完全信任大型的复杂项目。4.8 Conclusion This chapter has considered the emergence and growth of SLA within the framework of opposing views of language. That is not the same as a survey of SLA in its totality, nor does it claim to be. What I have tried to demonstrate in this chapter is that, even in SLA, which perhaps more than other sub-disciplines within applied linguistics often lays claim to be scientific, the paradigms within which researchers conduct their studies and debates are infused with quite opposite basic views of what constitutes relevant language data. Not to ask the question of what should constitute data for understanding SLA, and not to refer that question to basic ideological issues of what language is and what the object of study of linguistics should be sits badly with our notions of bi- directional accountability to theory and practice that we outlined in Chapter 1. A complete theory of SLA of course embraces more than this one chapter can hope to achieve. Issues not directly addressed in this chapter also play a major role in how SLA theories are shaped and changed, such as immersion-learning (e.g. Cohen and Swain 1976;Cummins and Swain 1986;Genesee 1987;Swain and Lapkin 1990), content-based language instruction(Mohan 1985), the study of bilingualism and multilingualism (e.g.Cummins 1991;Schreuder and Weltens 1993), the understanding of real-world cognitively demanding abilities such as competence in translation and interpretation(Schweda and Nicholson 1995), and the effectiveness or otherwise of methods of testing and assessment in relation to what they can tell us about SLA. This last issue is by no means a simple matter of efficient and objective metrics: social and cultural factors and the ideological aspects of testing contexts raise very fundamental questions about the use of testing as a measure of second language acquisition (Shohamy 1997;Larsen-Freeman 1997:90), and alternative methods of testing may fit better with socio-culturally oriented approaches to SLA(see Hamayan 1995 for an interesting discussion). What is more, a complete theory of SLA would need to account not only for language acquisition but also fo