by Rod Ellis
Given that the acquisition of implicit knowledge of a grammatical feature is a slow and dynamic process, it is difficult to see how short-term, explicit interventions can have anything but a superficial effect on the acquisition of entirely new features. It was this that led Krashen (1982) to support the non-interface position and propose the virtual abandonment of explicit instruction. However, research has shown that explicit instruction is effective in helping learners gain more control over grammatical features that they have already partially acquired. Also, explicit instruction may have a long-term indirect effect. According to the weak-interface position, instruction facilitates rather than causes acquisition. That is, the explicit knowledge that results from explicit instruction can assist the gradual development of implicit knowledge by facilitating noticing and noticing-the-gap when the learner subsequently experiences input. In other words—even if explicit instruction does not have an immediate effect on the acquisition of implicit knowledge—it can have an indirect, facilitative effect.
Implicit instruction makes use of ‘tasks’ to provide opportunities for meaning-centred interaction in the classroom. Through careful task-design and choice of implementation options, research has shown that it is possible to vary the extent to which learners attend to complexity, accuracy, or fluency when they perform different tasks. Task-based teaching also affords opportunities for a focus-on-form: that is learners are encouraged to pay attention to specific linguistic forms while they are communicating. This can be achieved in a variety of ways but in particular through corrective feedback. As noted earlier, corrective feedback in the context of meaning-focused communication can assist acquisition. It has also been shown to help learners progress along developmental sequences.
The relative merits of explicit and implicit instruction continue to be debated—not least because they entail different theoretical positions about L2 acquisition. The non-interface position has few supporters these days: the debate centres on the strong and weak-interface positions. Can explicit instruction lead directly to the knowledge needed for accurate use of the target feature in spontaneous language use? There is evidence that it can, but this might be because it results in automatized explicit knowledge rather than true implicit knowledge. However, if there is no functional difference between these two types of knowledge—as DeKeyser (2003) argues—then, it does not matter as explicit instruction clearly ‘works’. Implicit instruction is also effective. One advantage of implicit instruction is that it not only assists the development of linguistic competence—i.e. implicit knowledge—but also fosters interactional competence—for example, how to initiate and manage conversations in the L2.
The research investigating explicit and implicit instruction also offers teachers some useful tools for teaching grammar. Concept-based Language Instruction, Processing Instruction, Consciousness-raising Instruction and Task-based Language Teaching are all fully-theorized approaches that also offer the practical means for implementing them. The study of explicit and implicit instruction has not only increased our understanding of L2 acquisition, but has also enriched language pedagogy.
Applying SLA
I have argued that SLA constitutes an academic discipline in its own right (see Table 12.1). That said, the origins of SLA lay in a wish to improve language pedagogy—just about all the early figures in SLA were at one time teachers or language teacher educators. Research in SLA has continued to be driven by these same motivations, leading to discussion about how SLA can best serve the needs of language pedagogy. This is not an easy issue to address, for—as Bardovi-Harlig (1995) noted—the ‘relationship of pedagogy to second language acquisition is a complex one that is not clearly agreed on by applied linguists’ (p. 151). I will consider a number of different ways of making use of SLA for language pedagogy.
Basing pedagogical proposals on the findings of SLA
For some researchers, SLA provides ‘hard evidence’ which should be used to advise teachers about what approaches and techniques work best (Long 1990). In general, however, SLA researchers have been wary of prescribing or proscribing how to teach, preferring instead to exercise caution about applying SLA findings.
One way in which researchers have tried to make the results of their research of practical value to teachers is by an ‘implications’ section tacked on to the end of a research report. There is, however, an obvious danger in such an approach. It does not follow that the implications drawn from a single study are of relevance to all teachers in all instructional contexts. It is also doubtful whether teachers or, indeed, many teacher educators read research articles. so their ‘implications’ are likely to have little impact on language pedagogy.
A more promising approach, perhaps, is to base advice on a theory of L2 acquisition that has been tried and tested through research. This is the approach that Krashen adopted. His theory—the Input Hypothesis—has had a considerable influence on language pedagogy. It is frequently referred to in popular guides for language teachers (e.g. Ur 1996; Hedge 2000) and many teachers have heard of him. However, Krashen’s theory is no longer accepted in its entirety and no other single theory has replaced it as a guide for pedagogy. We are currently in a situation where there is a plethora of SLA theories offering different—and often conflicting—accounts of L2 acquisition. Thus, while there might be merit in exploring the applications of individual theories to language pedagogy, it would be asking a lot of teachers and language educators to find their way through the varying applications of the different theories, especially as much of the SLA literature is highly technical and so not readily accessible to lay readers.
A much better way is that adopted by Lightbown (2000). She proposed a set of generalizations which were ‘consistent with the research to date’ and which could serve as a ‘source of information which could help teachers set appropriate expectations for themselves and their students’ (2000: 431). Examples of her generalizations are:
There are predictable sequences in L2 acquisition such that certain structures have to be acquired before others can be integrated.
Knowing a language rule does not mean one will be able to use it in communicative interaction.
For most adult learners, acquisition stops … before the learner has achieved native-like mastery of the target language.
These generalizations were formulated with teachers in mind. Wisely, Lightbown did not propose specific pedagogical practices based on them. She saw SLA research as a body of knowledge that can help to shape teachers’ ‘expectations’ rather than as a source of specific recommendations. She called for researchers to ‘enter into a dialogue with classroom teachers’, listening to what they are saying as well as informing them about SLA. However, there is a problem with such generalizations. They are challengeable, especially as new research becomes available. We have seen—for example—that not all SLA researchers currently accept the idea of ‘predictable sequences’ or the idea that ‘acquisition stops’ for learners.
Practitioner research and SLA
A second way of making use of the findings of SLA is to encourage teachers themselves to participate in investigating L2 learning in their own classrooms. There are two ways of going about this. One is to carry out small-scale action research projects to investigate specific problems that teachers have identified in their own teaching (see Burns 2010). For example, teachers may be uncertain about how to go about correcting their students’ errors in speaking activities, so they could consult the SLA literature on corrective feedback and then design a study to examine whether a particular type of corrective feedback—for example, elicitation—works for them.
The second way is to encourage teachers to replicate an existing SLA study. Vasquez and Harvey (2010) demonstrated the value of this. They asked a group of MA and doctoral students taking a course in SLA to replicate Lyster and Ranta’s (1997) study of corrective feedback. To help them, they broke down the research process into a number of steps. First, they asked them to video record their o
wn classes, then to prepare transcripts of the lessons, and then—after extensive discussion of Lyster and Ranta’s categories—to code the data. Vasquez and Harvey reported that the teachers were surprised to find how prevalent recasts were in their own teaching and became more inclined to acknowledge the cognitive rather than the affective dimension of corrective feedback as a result of their research. Replicating SLA studies in this way can help teachers to examine their own beliefs about teaching and their actual practices.
Practitioner research provides an excellent way of making the link between SLA and teaching. However, it is time-consuming and probably not something that most teachers are likely to undertake regularly. One way that might appeal to teachers is small-scale investigations of particular instructional activities. For example, they could carry out an evaluation of an information-gap task to see whether it provokes negotiation of meaning and modified output when performed by their own students.
Exploring language pedagogy through SLA
A third way of making use of SLA for language pedagogy is to take pedagogic issues as the starting point and then scrutinize them from the perspective of SLA. Ellis and Shintani (2013) adopted this approach. They examined a number of popular teacher guides—ones that figure frequently in teacher education programmes of various kinds—and extracted from them a number of key pedagogic topics, such as how to teach grammar and vocabulary; the importance of ‘authenticity’ in teaching materials; error correction; and the use of the students’ L1 in the classroom. Such topics constitute ‘persistent concerns in the professional practice of teachers’ (Hedge 2000: 1). Ellis and Shintani first reviewed the pedagogic proposals about these issues in the guides before turning to the findings of SLA relevant to each issue. In other words, they did not so much ‘apply’ SLA as explore the validity of the proposals that figured in the guides.
One example of how this approach can help to illuminate language pedagogy will have to suffice. The guides typically advise teachers not to engage in corrective feedback while students are engaged in performing communicative tasks on the grounds that this will detract from the main purpose of the tasks, namely to promote ‘fluency’. The corrective feedback research in SLA, however, indicates a clear advantage in correcting errors while learners are primarily focused on meaning (see Chapter 6 and Chapter 11). The research also shows that corrective feedback need not interfere unduly with fluency, especially if it takes the form of recasts or elicitation. It would seem, then, that—in this case—there is a mismatch between what is promoted as ‘good practice’ in the teacher guides and what SLA has demonstrated.
Many teacher education programmes include a course on SLA. Perhaps the time has come to rethink how SLA is incorporated into such programmes. Relevance and transfer of knowledge may be better achieved by identifying interface issues—i.e. issues like corrective feedback that are relevance to both pedagogy and SLA—and then examining the extent to which current pedagogic views are supported or challenged by SLA research. SLA is—of course—not the only source of knowledge of relevance to language pedagogy, but it is surely an important one if language pedagogy is to ensure teaching-for-learning.
Conclusion
I concluded the first edition of Understanding Second Language Acquisition with the comment ‘there is no consensus about the overall direction that SLA research should follow’ (p. 290). This is perhaps even more clearly the case today. If anything, the epistemological differences that divide cognitive and social SLA—both with regard to theoretical positioning and to research methodology—are greater than ever. Ortega (2011) commented that there is ‘unheeded and unprecedented theoretical diversity in SLA’ (p. 176). For some SLA researchers, this is a problem. To my mind, however—given the complexity of L2 acquisition—it is both inevitable and desirable. Debate is healthy, and I do not know of any discipline in the social and psychological sciences where similar debate is not to be found.
Notes
1 Zhang and Lantolf (2014) report a study that shows that instruction can override the natural tendency of learners to acquire grammatical structures in a fixed sequence. To some extent at least, the route that learners follow is teachable.
2 Birdsong’s (2006) review of the research concluded that there was no sharp discontinuity, but rather a gradual decline in ability, starting from an early age.
Glossary
Acculturation Model This theory treats L2 acquisition as one aspect of acculturation (the process by which the learner becomes adapted to a new culture). Various factors influence the social and psychological ‘distance’ of the learner from the target-language culture and thereby the rate and ultimate success of L2 acquisition. See also social distance and psychological distance.
accuracy Accuracy ‘concerns the extent to which the language produced conforms to target language norms’ (Skehan 1996: 22). A typical measure of accuracy is percentage of error-free clauses.
accuracy order This refers to the order in which different grammatical features of an L2 reach native-like accuracy. See also morpheme studies.
Adaptive Control of Thought (ACT) Model This is Anderson’s model of skill-learning. The model accounts for how learners’ ability to perform a skill develops from a declarative stage, where information is stored as facts, to an automatic stage, where information is stored as easily accessed procedures.
Aspect Hypothesis The Aspect Hypothesis claims that acquisition of verbal morphology is determined by the lexical aspectual class of the verb i.e. whether the verb refers to a state—for example, ‘seem’ and ‘know’—an achievement —for example, ‘arrive’ and ‘fall asleep’—, an activity—for example, ‘sleep’ and ‘study’—or an accomplishment—for example, ‘build a house’ or ‘paint a picture’.
attention In SLA, attention is the cognitive process by which learners perceive and rehearse L2 features in working memory. The extent to which attention is an entirely conscious process or is subconscious remains a matter of some controversy. See also Noticing Hypothesis.
attractor state This is a term used in Dynamic Systems Theory. De Bot and Larsen-Freeman (2011) defined it as ‘the state the system prefers to be in over other states at a particular point in time’ (p. 14). Such states are not permanent but depending on the strength of the attraction may resist change for a period of time.
Attribution Theory The term ‘attributions’ refers to the subjective reasons by which we explain our past successes and failures. These can have a positive or negative effect on our motivational disposition.
Audiolingual Method A method developed in the United States based on behaviourist learning theory. It emphasizes the importance of pattern practice through mimicry and repetition in order to develop correct ‘habits’.
avoidance Avoidance is said to take place when specific target-language features are under-represented in the learner’s production in comparison to native-speaker production. Learners are likely to avoid structures they find difficult as a result of differences between their native language and the target language.
basic variety This constitutes an early stage of L2 acquisition identified by researchers in the European Science Foundation Project. It is characterized by the absence of grammatical functors. Learners rely instead on pragmatic means to convey semantic concepts such as pastness. Verbs are non-finite. The ‘basic variety’ is preceded by the ‘pre-basic variety’—characterized by nominal organization—and is followed by the ‘post-basic variety’—where finite verb forms appear.
behaviourist learning theory Behaviourist learning theory is a general theory of learning—i.e. it applies to all kinds of learning. It views learning as the formation of habits when the learner is confronted with specific stimuli which lead to responses, which are, in turn, reinforced by rewards, or are corrected. Behaviourist learning theory emphasizes environmental factors as opposed to internal, mental factors.
blocking See overshadowing.
careful style This is a term used by Labov (1970) to refer t
o the language forms evident in speech that has been consciously attended to and monitored. A careful style is used in formal language tasks such as reading pairs of words or doing a grammar test. See also vernacular style and stylistic continuum.
Cognition Hypothesis This constitutes a framework developed by Robinson (2001) for classifying the design and implementation variables that influence the complexity, accuracy, and fluency of L2 production.
cognitive comparison This is a term used to refer to the mental process involved when learners compare their own output with the input made available to them and identify the differences. See also noticing-the-gap.
communication strategy A communication strategy is employed when learners are faced with the task of communicating a meaning for which they lack the requisite linguistic knowledge—for example, when they have to refer to some object without knowing the L2 word.