Second Language Acquisition Research
The relevance of the fourth influential tradition in applied linguistic work, second language acquisition research or SLA, has been already been dealt with in some detail in Chap. 3 (“The maturity of applied linguistics”) above, while in Chap. 4 there is an analysis of early conceptions of the role of a psychology of language in applied linguistics as well as the divergent conceptions of SLA, as constituting either a separate discipline, or one more or less directly relevant to language teaching. A recent able review of the main issues in the diverging ‘practical’ (or “second language learning”) and ‘theoretical’ views of SLA can be found in Davies’s (2013: 36-51) discussion of these in the broader context of the ‘native’ speaker and user debate.
There is little doubt that those early studies, for example on the discovery of a purportedly invariant order of acquisition of morphemes and even of syntactic structures (Clark and Clark 1977: 342ff.; for some subsequent debates see Truscott 1996: 337; Ellis and Larsen-Freeman 2006: 558-559) received their initial stimulus from a Chomskyan perspective on what makes language development possible. It is clearly a mentalist starting point that encourages researchers to assume that, if humans are genetically predisposed to learning language, the activation of those principles of mind are common across individuals. The order in which structures are learned is a fixed order because in the rationalist perception the genetically in-built ability, the mechanism of the cognitive organization of language, is the same for all humans.
The relevance for language instruction of such an order of acquisition is immediately evident, but two further assumptions - that have always been contestable as well as contested - need then to be made. The first is that first and second, or first and subsequent language development would be similar, or at least, if the order of acquisition is shown to be different, that there would in the learning of an additional language still be a discoverable order, that would in its turn be relevant to second or additional language instruction. Just how quickly this first assumption became contestable, even for first language acquisition, is evident already in the early work done by Hatch (1978), Ritchie (1978) and others (e.g. Cook 1981a, b). Hatch (1977: 6), Henning (1977) looks at explanations for the discovery of a fixed order of learning among both first and second language learners not from the perspective of genetically in-built mechanisms of mind, but from the angle of how the structure of the discourse influences interaction among lingual subjects. In the latter explanation, the organization of the discourse provides slots into which fit the words and morphemes that have been identified. If that is the case, language emerges not from the mind, but from the interaction: if there is an invariant order, it therefore has a social rather than a genetic origin. The debate has endured in different forms for more than 20 years, and right up to the present: several of the contributions in Lee et al. (2009), for example, continue to make the point, but now from a dynamic systems theory perspective, that the emergence of language is best explained from the point of view of the empirical given of interaction through language, rather than from a rationalist and generativist perspective. Where in the latter view competence is idealized and separated from what are thought to be the stops and starts, idiosyncrasies and imperfections of lingual performance, in the perspective of Lee et al. (2009: 8) “... performance is competence”!
The second contestation derives from doubts about the actual practical relevance of the discovery of an order for learning. In the first instance not everything that is identifiable as having been learned, or that can potentially be learned, is teachable.
Instruction cannot to be equated with learning. Moreover, even though the initial order may be similar up to a certain point, how far down the list of forms to be acquired does the ‘invariability’ hold before individual differences start making generalizations impossible? All of this therefore potentially detracts from the proposition that SLA discoveries may be immediately and automatically relevant for instructional design.
The purportedly Chomskyan origins of SLA are, however, not the whole story, historically speaking, about this fourth paradigm of applied linguistic endeavor. The sociolingual explanations offered by Hatch (1977, 1978) and others for language learning within the field of SLA, and referred to above, provide a good illustration of a broader than formal linguistic basis for this kind of research. One should thus question Spolsky’s (2005 : 28f.) claim that second language acquisition research derives merely from a
... misguided ... approach (that) involved accepting Chomsky’s attack on Skinner as weakening all claims for grammar learning, and adopted Chomsky’s use of the term ‘acquisition’ and his search for universal innate grammars in order to condemn all effort to teach languages, glorifying automatic acquisition through exposure to the next needed form . This had major effects on the profession, justifying the building of a new sub-field to study ‘second language acquisition’ .
This misrepresents the importance of fourth generation work in applied linguistics, in that this style of working within the discipline provided a first psychological justification for the most important development in language teaching in the late twentieth century, communicative language teaching (CLT) : As was the case with its linguistic justification, it did so belatedly, after the fact, but it was nonetheless important as a theoretical rationale for the design not only of language teaching, but also of language assessment (Green 2014: 198ff.). It is important to note, furthermore, that this theoretical justification within SLA studies in general was not limited to a single, generativist paradigm. There is within SLA studies a broad a range of perspectives on what constitutes and also on what influences language development.
The quest for SLA studies relevant to the design of language instruction originates in a line of research developed by Pica and her associates, as well as by other, sometimes more critical voices, from the mid-1980s onwards (Pica and Doughty 1985 ; Gass and Madden 1985 ; Varonis and Gass 1985; Doughty and Pica 1986; Porter 1986; Pica 1987,1994,1996; Pica et al. 1987,1989,1993; Walz 1996; Foster 1998; Oliver 2000; for another summary and review of numerous other studies, cf. Nunan 1991). The relevance of this approach has already been referred to, but probably finds its best articulation in the ongoing work of Lightbown and Spada (2006; cf. also Larsen-Freeman 1993).
Because this line of research often focuses on how meaning is negotiated in interaction with others (most often in bridging an information gap), thus providing evidence of potential learning gains that derive from communication amongst peers, the psychological justification for CLT that is to be found in this kind of acquisition studies is closely related also to the next tradition in applied linguistics that will be discussed here, namely constructivism. This kind of SLA research, representative of what is being called fourth generation applied linguistics in the current analysis, thus also shows how continuity between and among different styles of applied linguistics is the norm rather than the exception. Each has a distinct characteristic, yet many connect backwards and forwards with other traditions and styles of designing language interventions . In fact, in the later work of Dornyei (2009, 2010; cf. too Murphy 2010) there is a clear indication that SLA research, even where it examines individual learner differences from social and educational points of view, has turned towards the adoption of a dynamic systems theory (DST) paradigm. the postpostmodernist challenger that will be discussed below.
As has already been observed, the regular and periodic reviews of the findings of this kind of research for second language teaching (Tarone et al. 1976; Cook 1978; Burt and Dulay 1981; Dulay et al. 1982; Lightbown 1985; Spada 1997; Lightbown 2000; Lightbown and Spada 2006; Ellis and Larsen-Freeman 2006) invariably emphasize that their insights need to be treated rather as implications than as applications . In the same manner, Spolsky (2008: 5) refers in a discussion of potential psycholinguistic and neurobiological contributions to “the realization that the core fields do not have direct application but rather set possibilities and have implications for activity.” Referring to numerous examples from his own experience, as well as to investigations of how actual language learning processes happen, Nunan (2000: 7) has correctly observed that “the relationship between teaching and learning is asymmetrical - in short ... learners do not learn what teachers teach in a iinear. additive fashion.” This echoes an earlier finding by Lightbown (1985: 181) that a model of learning that assumes that language can, like building blocks, steadily be built up by adding bits of grammar, is a wholly unrealistic conception with no support in the research.