by Rod Ellis
As Sfard (1998) argued, both the ‘participation’ and ‘acquisition’ metaphors are needed along with the recognition that it is not possible to make a clear distinction between the two. Cognitive SLA is arguably guilty of under-theorizing the nature and importance of the social context. Some social turn researchers are guilty of failing to theorize what acquisition entails and struggle to handle the cognitive side of L2 learning. As Veronique (2013) argued ‘an integrated research perspective is highly desirable’ (p. 270)—one that combines social, psychological, and linguistic perspectives and gives weight to all three. At present, though, there is little sign of this happening. As Firth and Wagner (2007) observed when they revisited the arguments in their 1997 paper in support of a social dimension in SLA, the field may not be able to ‘withstand the current bifurcations, competing methods, critiques, and internal tensions, and remain generally cohesive’ (p. 813).
Notes
1 Block (2003) is clearly right to point out the problem with the term ‘second’ as, for many learners ‘additional’ is clearly a more appropriate term. However, ‘second’ continues to be the preferred term. For example, the title of Atkinson’s (2011a) book, which provides an account of alternative, more socially-oriented approaches, refers to ‘second’ language acquisition. Once established, labels are not easily replaced.
2 Lantolf (2006) suggested that when even advanced L2 learners are faced with a difficult task, they have difficulty in sustaining use of the L2 and are likely to resort to the use of their L1 in their private speech.
3 Swain’s research demonstrates that both collaborative dialogue and self-verbalization can mediate learning. It would be interesting to know if one is more effective than the other. Lantolf (2011), however, noted that this interesting question remains unresearched.
4 However, CA-SLA has recently begun to acknowledge that Discursive Psychology might have a role to play. Discursive Pyschology acknowledges that there are psychological phenomena, but sees these as things that are constructed, attended to, and understood in interaction.
5 Sociocultural researchers have not been so reluctant to avoid testing. Also, see Ellis and Shintani (2013) for a study that combined a conversation analytic approach by tracking learning objects with pre- and post-tests.
6 Norton Peirce (1995) argues that ‘investment’ is not the same as either ‘integrative’ or ‘instrumental motivation. However, as she describes it has clear connections with Dörnyei’s (2005) ‘ideal self’ in his L2 Motivational Self System (see Chapter 3).
7 In this respect, the theories of the ‘social turn’ view language in a very similar way to the theories that see language as a complex adaptive system, which can also lay claim to being ‘social’ in their epistemology (see Chapter 8).
10
The role of explicit instruction
Introduction
All language instruction constitutes a form of intervention in the process of learning a second language (L2). In this respect, instructed language learning differs from the naturalistic language learning that takes place in first language (L1) acquisition and in untutored L2 acquisition. However, it does not follow that contextual differences are necessarily reflected in differences in the cognitive and social processes involved in acquisition. Indeed, what is of interest to SLA researchers is precisely whether instruction does involve different learning processes. Investigating the effects of instruction, therefore, serves as a means of testing different claims about the nature of the processes involved in L2 acquisition. In this and the next chapter, I will examine this research. My main purpose is to use the research to examine some of the competing theories discussed in previous chapters in order to further understanding about the nature of L2 acquisition. In so doing, however, I will also address a question of paramount importance to teachers—namely, how can instruction best foster language learning?
Types of intervention
Instruction as intervention in the process of language learning is of two basic kinds—direct and indirect. Direct instruction involves providing learners with explicit information about the target of the instruction, often together with opportunities to practise the target. I will henceforth refer to this type of instruction as explicit instruction. Explicit instruction invites intentional learning on the part of the learner. Indirect instruction involves setting up opportunities for learners to learn without specifying what the target of the instruction is: in other words, there is no provision of explicit information about the target feature, although there are opportunities to engage in the use of it. I will refer to this second type of instruction as implicit instruction. It caters to incidental learning on the part of the learner, although learners may sometimes choose to focus on specific linguistic forms and try to learn them intentionally. Table 10.1 provides a more detailed specification of these two types of instruction (see also Ellis and Shintani 2013: Chapter 4).
Implicit instruction Explicit instruction
attracts attention to target form
directs attention to target form
is delivered spontaneously (e.g. in an otherwise communication-oriented activity)
is predetermined and planned (e.g. as the main focus and goal of a teaching activity
is unobtrusive (minimal interruption of communication of meaning)
is obtrusive (interruption of communicative meaning)
presents target forms in context
presents target forms in isolation
makes no use of metalanguage
uses metalinguistic terminology (e.g. rule explanation)
encourages free use of the target form.
involves controlled practice of the target form.
Table 10.1 Implicit and explicit forms of form-focused instruction (based on Housen and Pierrard 2006: 10)
There is also another distinction commonly referred to in the literature. Focus-on-form refers to an approach that involves an attempt to draw learners’ attention to a linguistic form while they are primarily focused on meaning—i.e. trying to communicate. It contrasts with focus-on-forms—seen as the ‘traditional approach’—in which the primary goal is to help learners master the structural features listed in the syllabus by making the linguistic target of each lesson quite explicit. Here ‘the aim is to direct learner attention and to exploit pedagogical grammar’ (Doughty and Williams 1998: 232). These definitions suggest that focus-on form involves implicit instruction and focus-on-forms explicit instruction. This is broadly the case. However, these distinctions are not the same. Implicit instruction does not necessarily involve a primary focus on meaning. Some researchers have investigated implicit instruction by simply telling students to read and memorize a set of discrete sentences—an activity that, arguably, does not require a primary focus on meaning. Also, explicit instruction can include practice activities that require learners to focus on meaning. The terms ‘focus-on-form’ and ‘focus-on-forms’ are perhaps best used to distinguish specific kinds of instructional activities that may figure in both implicit and explicit instruction.
The focus of this chapter is on explicit instruction and the following chapter examines research that has investigated implicit instruction. Where appropriate, I will use the terms ‘focus-on-forms’ and ‘focus-on-form’ to refer to specific activities that figure in studies that have investigated these two types of instruction.
Types of explicit instruction
Explicit instruction is not a uniform approach. It can be provided in a number of different ways, involving different theoretical assumptions concerning the role of explicit information and different kinds of practice activities. Table 10.2 outlines the different approaches along with the theoretical positions that support them.
General types of explicit instruction Instructional approaches Interface position Theoretical basis
Deductive pedagogic grammar and production practice (as in presentation-practice-production, i.e. PPP) strong Skill-learning theory (DeKeyser 1998)
Deductiv
e integrated instruction (i.e. explicit explanation provided during communicative practice) strong Transfer Appropriate Processing (Lightbown 2008)
Deductive concept-based instruction (involving presentation of ‘scientific concepts’ and production practice strong Sociocultural theory (Lantolf and Thorne 2006)
Deductive comprehension-based instruction (as in Processing Instruction) strong Input Processing Theory (VanPatten 1996)
Inductive pattern practice (as in the Audiolingual Method) strong Behaviourism (habit formation)
Inductive consciousness-raising instruction (involving consciousness-raising tasks) weak Theory of Instructed Language Learning (R. Ellis 1994)
Table 10.2 Different kinds of explicit instruction
A general distinction can be made between deductive and inductive instruction. In deductive instruction, learners are provided with metalinguistic information about the target of the instruction. This can occur in the presentation stage of the lesson or while learners are engaging in practice activities. Deductive instruction also differs in terms of the nature of the explicit information provided. Generally, the explicit information is typically drawn from a pedagogic grammar—i.e. it is simplified—but, in some types of explicit instruction, it consists of ‘scientific concepts’—i.e. very detailed linguistic descriptions of grammatical features. In inductive explicit instruction, no metalinguistic information is provided. In the Audiolingual Method, for example, controlled production practice is employed to develop correct ‘habits’, whilst in Consciousness-Raising Instruction, learners complete a series of tasks that guide them to an understanding of the target feature, but without being asked to practise it. Both deductive and inductive instruction also vary in terms of whether the practice activities involve production or comprehension of the target feature. Later, we will consider different types of explicit instruction that incorporate these various options.
In Chapter 1, I outlined three different positions relating to the relationship between explicit and implicit knowledge (see also Chapter 8). The different instructional approaches listed in Table 10.2 reflect two of these positions: the strong interface position, which views explicit knowledge as transformable into implicit knowledge (or its equivalent), and the weak interface position, which views explicit knowledge as a facilitator of implicit knowledge but not directly transformable into itNOTE 1. Investigating these instructional approaches, then, can help to shed light on the legitimacy of these two positions. Each of the instructional approaches draws on a different theoretical base: for example, whereas presentation-practice-production (PPP) draws on skill-learning theory and emphasizes the importance of production practice, Processing Instruction draws on VanPatten’s Input Processing Principles and emphasizes the need for comprehension-based practice. Thus, investigating studies that have compared these two instructional approaches enables the claims of the theories that support them to be evaluated. In the sections that follow, I will review some of the key studies that have investigated these instructional approaches and their theoretical underpinnings.
Presentation-Practice-Production instruction
Presentation-practice-production (PPP) constitutes the dominant approach promoted in teacher guides (e.g. Scrivener 2005; Ur 1996). As noted in Chapter 9, it is supported by Skill-Learning Theory, which claims that language learning is like any other kind of learning. The starting point is declarative knowledge (i.e. explicit knowledge of specific linguistic forms), which is then proceduralized through practice until eventually it becomes automatic. Acquisition of linguistic forms, therefore, is conceptualized as involving a shift from controlled to automatic processing. The role of instruction is to facilitate this process by first providing learners with explicit knowledge of a target feature and then, through practice, facilitating the cognitive changes needed for automatic processing. In effect, then, PPP assumes a strong interface between explicit and implicit knowledge although, as DeKeyser (2003) pointed out, the end result may not be ‘true’ implicit knowledge but rather speeded-up declarative knowledge, which he argued was functionally equivalent. The extent to which even this takes place, however, is disputed with some researchers (e.g. Hulstijn 2002; Paradis 2009) claiming that there are limits to how automatized explicit knowledge can become and that it is not equivalent to implicit knowledge.
The test of skill-learning theory is whether PPP results in the learners’ ability to communicate freely using the language feature targeted by the instruction. To investigate this, I have chosen a number of carefully designed studies whose instructional components reflect the principles of PPP. That is, the explicit component should provide a clear explanation of the target feature and the practice component should include a variety of activities including communicative tasks. The inclusion of communicative tasks is seen as of special importance because—as DeKeyser (1998) emphasized—for automatization to happen, learners need to practise the target feature under ‘real operating conditions’. Three further criteria for the selection of the studies were: (1) they included a control group that did not receive the instruction; (2) the testing regime included a measurement of learning based on free production (i.e. not just discrete point tests); and (3) there was a delayed test to provide some evidence of long-term learning. I have also selected studies that investigated a variety of linguistic features—grammatical, lexical, sociopragmatic, and phonological. Table 10.3 summarizes six studies that meet these specifications.
Study Target feature Participants Instruction Tests Results
Harley (1989) two French verb tenses—passé composé and imparfait 319 grade-six students in a French immersion programme in Canada explicit description of the target structures followed by eight weeks of a mixture of practice activities including communicative tasks ratings based on written compositions; cloze test scores; errors scores based on an oral interview. The instructed learners outperformed the control group on all measures derived from both the immediate and delayed post-tests—including those based on the oral interview.
Lyster (1994) French address forms tu and vous 106 grade-eight students in a French immersion programme in Canada explicit techniques (e.g. comparing use of tu and vous); structured exercises; intensive reading activities, free writing, role plays. an oral production test, a written production test, and a multiple-choice test. Immediate and delayed tests showed clear gains in both oral and written production and also awareness of socio-stylistic differences.
Housen, Pierrard, and van Daele (2006) 69 14–15-year-old Dutch speaking learners of L2 French in Belgium French sentence negation (a ‘simple structure’); French passive constructions (a ‘complex structure’) four weeks of instruction consisting of: (1) pedagogical rule; (2) reading text; (3) identification of exemplars in the text; (4) description of these examples; (5) sentence-transformation and answering semi-open questions. an unplanned production task consisting of oral questions about pictures and objects, which the learners had to answer instantly. No difference between experimental and control groups on absolute number of tokens of target structure produced on the immediate post-test, but experimental group significantly outperformed the control group in accuracy of production. Same pattern for delayed post-test. The complexity of the target structure did not affect the results.
Akakura (2012) Non-generic and generic uses of English articles (a and the) 94 ESL advanced-level students in a private language school in New Zealand—mainly Asian computer-based; explicit explanation of functions performed by articles; controlled and free production activities. an elicited imitation task and an oral production task to measure implicit knowledge; a grammaticality judgement task and a metalinguistic knowledge test to measure explicit knowledge (see R. Ellis 2005c) The instructed learners outperformed the control group in the immediate and delayed metalinguistic knowledge test and elicited imitation task. However, no effect was evident for the generic use of articles in the oral production task. Effects of instruction stronger in delayed tests.
Shintani (201
1) 24 nouns and 12 adjectives 30 complete beginner Japanese learners of English as a foreign language explicit presentation of the target words followed by controlled and free practice activities. discrete word production test; same or different task The PPP group outperformed the control group. Significant gains for both nouns and adjectives evident in both tests. The gains were maintained in the delayed post-test.
Saito and Lyster (2012) English /r/ as in ‘red’ 65 intermediate-level Japanese learners of English in Japan Four hours of form-focused instruction integrated into argumentative skills lessons containing 38 minimally paired words. Instruction carried out with and without corrective feed back to two experimental groups. Control group received no instruction and no corrective feedback. ratings and acoustic analyses conducted on /r/ tokens of familiar and unfamiliar items elicited via a word reading test, a sentence reading test and a timed picture description task. The group that received corrective feedback demonstrated improvement in pronunciation of /r/ in both controlled and free production both immediately after instruction and later. In contrast, the group that received instruction without corrective feedback did not improve.