Input for Instructed L2 Learners: The Relevance of Relevance

Preface

This book is an attempt to apply relevance theory (RT) (Sperber & Wilson 1986/1995) to verbal input for instructed foreign language learners. First, I would like to define the scope of my discussion and my understanding of the terms used.

Input is difficult to define in the second/foreign (L2) classroom perspective¹ because, on the one hand, in its general sense, the term stems from information processing theory, where it denotes any verbal or non-verbal information that reaches one’s processing system, and on the other, in a more specific sense, the Comprehensible Input and Interaction Hypotheses have linked the concept of input in L2 learning and teaching with Second Language Acquisition (SLA) theory.²

However, SLA theorists and researchers who have tried to account for second language acquisition on the basis of an analysis of the linguistic data reaching one’s processing system have faced great problems in finding empirical support for the existence of specific input factors conducive to SLA (see Ellis, 1994). The reason for the problems seems to lie in the vagueness of the concept of input itself, as well as in the multiplicity of factors which affect successful language learning/acquisition (see Brown, 1994).

Being an L2 teacher and an L2 teacher trainer, I firmly believe that there is a link between teaching and learning/acquisition. Such a link is demonstrated in teaching and learning practice every day in thousands of L2 classrooms. On the other hand, I must admit that it is very difficult to find unequivocal evidence, conforming to a rigorous scientific paradigm, that some types of teaching, including some types of L2 classroom discourse, are more conducive to learning/acquisition than others.

Searching for innovative theoretical approaches to the aforementioned problems, we can begin our search from the teacher’s perspective, and her/his obvious intention to facilitate the process of L2 learning and acquisition. Thus, my intention is to treat input for instructed L2 learners in a different way. In my understanding of the term, ‘input for instructed L2 learners’ is not any verbal information that reaches the learners’ processing systems. It is ‘the language intentionally presented to the learners by the teacher or other learners in order to facilitate the process of L2 learning/acquisition’. Such an understanding of the term stems from the nature and basic goals of the L2 teaching process. I do not claim that the teacher’s input is always facilitative. I only say that the teacher wants the learners think that it is.

On this view, my intention is to conceptualise the teachers’ (or peers’) input within the framework of RT, which is a theory of the interpretation of incoming messages. The presentation of the L2 classroom input is understood as following the Principle of Relevance, that is, automatically communicating to the audience (the learners) a presumption of its optimal relevance (Sperber & Wilson, 1986/1995). By the above definition, I mean that the teacher, or the learner in the role of the teacher, according to the Principle of Relevance, makes their audience (the learners) believe that the input he/she provides is optimally relevant to them.

Interpretation of L2 classroom input has become an interesting issue in the light of RT, because the teachers’ intentions are not fully explicit. The impact of the Communicative Approach has contributed to considerable tensions within L2 classrooms, particularly those in foreign language learning contexts. Those tensions involve apparent conflicts between a focus on fluency and a focus on accuracy, and in monolingual contexts,³ additionally, between L2 and native language (L1) use. I would like to interpret those conflicts as stemming from a fundamental tension within the communicative L2 classroom, between a focus on communication and a focus on the target language code.

My intention is to analyse teachers’ (and peers’) input within L2 classroom discourse in the light of RT. However, before I do this, I would like to give an overview of two other closely related perspectives on L2 classroom input: an L2 teaching perspective and an interactional discourse analysis view.

My classroom data is based on a corpus collected in seven L2 classroom research projects by my former MA seminar students. All the classroom discourse data was collected in secondary school English as a Foreign Language (EFL) classrooms in Poland.

The first chapter presents an overview of the role of L2 classroom input in the light of SLA theory and its critique, followed by my main claims concerning the application of RT to classroom input for instructed L2 learners in a foreign language learning context.

The second chapter gives a teaching perspective on the role of the L2 classroom input in the communicative L2 classrooms. In particular, it focuses on fluency and accuracy practice, providing feedback and error correction, and L1 use in monolingual L2 classrooms.

The third chapter presents the L2 classroom discourse perspective on the role of input and interaction. The approaches involve a discussion on the differences between naturalistic and L2 classroom discourse, functions of L2 classroom discourse, patterns of participation in L2 classroom discourse, teacher talk and peer talk approaches to L2 classroom discourse and L2 classroom discourse modifications.

The findings of seven MA L2 classroom research projects from the years 1984 to 2004 are presented in their original versions in Chapter 4 to enable their reinterpretation in the light of relevance theory in the next chapter.

Chapter 5 analyses L2 classroom discourse samples according to the functional teaching categories: explicit teaching, including explicit presentation of the linguistic data and teacher corrections of learners’ language, as well as L2 classroom communication, subdivided into real communication and simulated communication.

In real communication, the analysis is focused on talking about the learning content, including subject-matter teaching, and talking about organisational and social matters. In simulated communication, two types of communicative activities are analysed: role-plays and discussions.

It is suggested that L2 classroom instructional input plays an important role in classroom discourse, because by changing the expected levels of relevance, it indicates to the learners how they should interpret it: as fluency practice, as accuracy practice, or as fluency combined with accuracy.

The final Chapter 6 presents teaching implications of the proposed interpretation of instructional input in the L2 classroom discourse, in particular for the development of fluency and accuracy in foreign language teaching contexts.

This book is first of all intended for L2 teacher educators, L2 teachers and pre-service and in-service teacher-trainees, in particular those working in the countries where English and other foreign languages are taught primarily in educational settings by non-native teachers.

I also believe that the book can be of interest to SLA researchers. In my opinion, a number of SLA researchers, who are predominantly linguists and have little to do with L2 classroom teaching, tend to disregard what actually happens in instructional settings, where the majority of students learn their L2, admittedly not to the level of near-native proficiency.

On the other hand, teachers and teacher educators, by virtue of their teaching focus, may overestimate the impact of instructional factors upon target language development. The two groups of professionals often work in different worlds although their research subjects remain the same – instructed L2 learners. My intention in this book is to link both perspectives on the grounds of RT.

Notes

1. The abbreviated term L2 refers to both second and foreign language. Second language is the language which is acquired/learned naturalistically and/or in the classroom, in the countries where it is spoken as a first language, e.g. English is a second language for non-native speakers in Great Britain. Foreign language is usually acquired/learned only in instructional settings in the countries where it is not spoken as a first language, e.g. English is a foreign language in Poland.

2. L2 acquisition is a term closely connected with SLA theory, stemming from a psycholinguistic claim that language learning (both first and second language) is first of all based on unconscious mental processes. On the other hand, L2 learning usually refers to intentional activities which aim at the development of the learners’ L2 knowledge. Frequently, both terms cannot be easily distinguished on theoretical grounds, and it is common to use the combined term L2 learning/acquisition. L2 teachers tend to avoid entering into the acquisition and/or learning dilemma, and they use another term L2 development, which denotes growing communicative competence in L2 use.

3. By monolingual classrooms I mean the classrooms where all L2 learners speak one common native language (L1). Most frequently, L2 teachers in monolingual classrooms are non-native L2 speakers and share the common L1 with their students.

Chapter 1

The Role of L2 Classroom Input in the Light of Second Language Acquisition Models and Relevance Theory

The Role of L2 Classroom Input in the Light of Second Language Acquisition (SLA) Models

Introduction

At first, it should be stressed that the understanding of the term input in SLA theory and in Relevance Theory is different. Although my aim is to view L2 input in terms of Relevance Theory, it is impossible to avoid references to L2 input as it is understood in SLA theory. That is why it seems appropriate to outline some current approaches to the role of input in SLA models.

Input is understood in them as raw (primary) L2 data (Gass, 1997) that reaches the non-native audience’s perceptual system, that is, the second language which is noticed by the audience. In terms of the L2 classroom, L2 input is the target language spoken by the teacher which is heard by the learners.

The SLA models which consider L2 input as one of the crucial factors in language acquisition, view the process of L2 comprehension as the decoding by the non-native audience of the meanings communicated by the native speakers. By the same token, the SLA models which are concerned with L2 classroom second language learning/acquisition view the L2 comprehension process as the decoding by the learners of the meanings communicated by the teachers.

On the other hand, Relevance Theory is concerned with the interpretation of the already decoded messages, and the main part of the comprehension process follows linguistic decoding. I will discuss the aforementioned difference in the following part of this chapter.

Krashen, Long and Swain (input/interaction/output models)

The first model which treated input, in the above raw L2 data sense of the word, as the main factor in L2 acquisition was Stephen Krashen’s Comprehensible Input Hypothesis (Krashen, 1981, 1982). Krashen claimed that the first necessary condition for the input to be acquired is its comprehensibility, which is in turn ensured by its approximate level of difficulty, slightly higher than the non-native speaker’s or the learner’s present proficiency level. Such input was called roughly tuned input.

The other necessary condition for the comprehensible input to be acquired, according to Krashen, is an accompanying low Affective Filter, which refers to the non-native speaker’s or the learner’s positive attitude towards L2 learning and everything that entails. According to Krashen’s model L2 acquisition will automatically occur when communication and comprehension are successful.

Critics of Krashen’s Comprehensible Input Hypothesis point out that his concepts are vague, e.g. it is not clear what is a slightly higher level of difficulty, and it is not explicitly stated whether they apply to all aspects and levels of L2 learning/acquisition (see McLaughlin, 1987). Moreover, Krashen’s theory conflates L2 acquisition and L2 comprehension, by claiming that once L2 input has been comprehended, it has also been automatically acquired, which, obviously, is not the case.

A similar SLA model, equating comprehension with acquisition of raw L2 input data was proposed by Michael Long (1983) as the Interaction Hypothesis. Long claimed that the input provided by native speakers for non-native speakers must be adjusted in interaction to become comprehensible. He identified a few types of interactional adjustments in conversations between native and non-native speakers, such as confirmation checks, clarification requests and comprehension checks. Long concluded that there exists an ‘indirect causal relationship between linguistic and conversational adjustments and SLA’ (Long, 1985: 388).

The indirect causal relationship was based upon deduction: if adjustments result in comprehension, and comprehension results in acquisition, then adjustments should result in acquisition. However, Long’s conclusion did not find consistent support in research studies, and even those studies that supported the hypothesis in its first part, concerning the relationship between adjustments and comprehension, did not support the second part, that comprehension equals acquisition (see Ellis, 1994).

In a weak version of the Interaction Hypothesis Long claims that the feedback on errors, received from the native speaker interlocutor during interaction can facilitate L2 development, but probably only in some aspects of L2 learning (Long, 1996), which is a much less radical claim, which could be much more easily accepted by L2 teachers.

In turn, Merrill Swain (1985) in a modification of the Comprehensible Input Hypothesis argued that comprehensible input alone, even in vast quantities, cannot make L2 learners fully competent target language speakers. What she postulated as a necessary condition for achieving native-like competence was ‘comprehensible output’, that is, the learner’s spoken language ‘as he or she attempts to create precisely and appropriately the meaning desired’ (Swain, 1985: 252).

Swain (1995) further elaborated her hypothesis, in which she distinguished three functions of output in L2 learning: the noticing function, the hypothesis-testing function and the metalinguistic function.¹ In the first and the third functions, learners’ output plays the role of input for them. According to Swain, learners’ own spoken language, that is, their output, helps them to notice gaps in their L2 knowledge and to reflect upon them. Consequently, learners’ output can function as input for conscious reflection.

However, in Swain’s output-as-input model, it is not clear how learners can notice gaps in their knowledge if they are not provided with any feedback on their errors. Moreover, even if they are aware of their deficiencies, they may not have time to reflect on them in oral communication.

Nevertheless, in contrast to Krashen’s and Long’s hypotheses, Swain puts stress on accuracy of target language forms. She realises that meaning-focused instruction does not suffice in acquiring accurate L2 forms.

Gass’s model

An integrated model was proposed by Susan Gass (1997) as an attempt at combining the Input/Interaction Hypotheses with the Universal Grammar Hypothesis² and cognitive approaches³ to L2 learning/acquisition. According to Gass, L2 input should be first noticed and related to the existing knowledge. Raw L2 data is claimed to be first filtered by a cognitive mechanism called apperception to become apperceived input.

Apperception is an internal cognitive act in which a linguistic form is related to some bit of existing knowledge (or gap in knowledge). We can think of apperception as a priming device that prepares the input for further analysis. Thus, apperceived input is that bit of language that is, noticed in some way by the learner because of some particular recognizable features. (Gass, 1997: 4)

The apperceived input is claimed to be understood due to the process of negotiation and input modification, which places Gass’s model as an elaboration of the Comprehensible Input and Interaction Hypotheses.

However, according to Gass, comprehended input is different from comprehensible input in that ‘the focus is on the hearer (the learner) and the extent to which he or she understands’ (Gass, 1997: 5). Gass also claims that in her model understanding may mean anything from general comprehension to detailed understanding of phonological or syntactic patterns.

At the next stage in Gass’s model comprehended input is either used for immediate communication and processed at the level of general meaning, or it is processed at the level of morphology, lexicon and syntax, and consequently incorporated into L2 learners’ grammar to become intake, that is, acquired language.

Gass’s psycholinguistic model seems oversimplified, particularly in that it makes a sharp distinction between general meaning-focused processing for immediate communication (for comprehension) and form-focused processing (for acquisition). Moreover, it is not clear what makes some comprehended input suitable for form-focused processing.

Ellis’s model

Another SLA model which places L2 input in the focus of attention and tries to integrate various aspects and stages of the process of L2 acquisition is the Integrated Theory of Instructed L2 Learning proposed by Ellis (1990). The author draws on Krashen’s and Long’s hypotheses, as well as on Bialystok’s and Bialystok and Sharwood-Smith’s cognitive models of L2 learning⁴ (Bialystok, 1978; Bialystok & Sharwood-Smith, 1985).

From the perspective of instructed L2 learners, Ellis’s model is interesting because it treats specific classroom input as an important factor in L2 acquisition. Ellis acknowledges that beside meaning-focused instruction, form-focused instruction has a role to play in L2 acquisition. He is aware of the complex nature of meaning-focused and form-focused instruction in the L2 classroom and the role played by the L2 teacher in shifting the focus from form to meaning and vice versa. In his interpretation, Ellis is close to the ideas expressed by Sperber and Wilson in Relevance Theory.

The input that derives from these two kinds of instruction [meaning-focused and form-focused] differs with regard to its communicative properties (e.g. meaning-focused instruction is likely to afford the learner an opportunity to listen to and to perform a greater range of linguistic functions than will form-focused instruction) and also with regard to the kind of response it typically evokes in the learner (e.g. form-focused instruction encourages the learner to reflect on the formal features of the language while meaning focused instruction encourages semantic processing […]. The instructional input in many lessons will be mixed, affording the learner the opportunity to attend to both meaning and form – to experience language or to study it. Teachers shift the focus as the lesson unravels – at one moment engaging the learners in meaningful communication and at another directing their attention to the linguistic code […]. Ultimately, of course, it is the learner and not the textbook or the teacher that determines in what way the input is attended to. (Ellis, 1990: 188)

The Integrated Theory of L2 Learning underlines the learners’ active role in attending to meaning-focused and/or form-focused input. According to Ellis, learners derive L2 explicit knowledge, i.e. knowledge about the language and conscious concepts, from form-focused instruction. Explicit knowledge is used to ‘sensitise the learner to the existence of non-standard form in her interlanguage and thus facilitates the acquisition of target-language forms’⁵ (Ellis, 1990: 195).

However, in Ellis’s view, it is only implicit knowledge, i.e. subconscious L2 knowledge, derived mostly from meaning-focused teaching, which is responsible for spontaneous L2 use. In other words, L2 is acquired first of all from meaning-focused teaching although form-focused instruction can somehow help the learners in the process.

The cognitive focus-on-form (FonF) approach

The cognitive focus-on-form approach (FonF) is an elaboration of the models which treat input to L2 learning as an important factor affecting cognitive processes of knowledge acquisition. In particular, the FonF model puts stress on L2 language forms as they are learned and/or acquired in communicative L2 classroom contexts. The approach draws on numerous research studies stimulated by the Interaction Hypothesis and its modification, the Output Hypothesis (Swain, 1985), which aimed at discovering input/interaction/output factors conducive to L2 acquisition (see Doughty & Williams, 1998).

It was the lack of native-like L2 acquisition in L2 classroom immersion contexts, where L2 learners had a great amount of comprehensible (and comprehended) input, which resulted in a renewed interest in focusing on L2 form in instructed L2 (Long, 1991).

What should be borne in mind is the distinction between focus-on-form (FonF) and focus-on-forms (FonFs) approaches. While ‘focus on form overtly draws students’ attention to linguistic elements as they arise incidentally in lessons whose overriding focus is on meaning or communication’ (Long, 1991: 45, in Doughty, 2001: 210), focus on forms is concerned with explicit teaching of L2 forms only.

According to Doughty (2001), cognitive processes that are involved in L2 learning can be divided into macroprocesses, one of them being acquisition of input, and microprocesses, such as focus on form.

The crucial question asked by FonF approach is whether and how a microprocess, such as focus on form, which consists of teacher intervention, can influence an inaccessible and automatic macroprocess, such as acquisition of input.

Doughty claims that L2 learners must notice the gap between their nonnative-like forms and the target forms made salient by the teacher. Such noticing is possible due to the cognitive resources L2 learners possess, one of which is ‘a cognitive preference for re-utilizing recent speech’ (Doughty, 2001: 229).

Doughty (2001: 249) makes an interesting case for the existence of ‘small cognitive windows of opportunity’, through which L2 teachers could intervene by focusing on form in otherwise meaning focused activities. Such an approach is close to my understanding of the L2 teacher’s role in the language classroom, to be discussed in the following part of this chapter and exemplified in Chapter 5.

The questions that are asked by FonF researchers concern the time when shifts of attention from meaning (fluency) to form (accuracy) should occur. In other words, when L2 teachers should intervene in communicative activities. Another equally important question refers to the form of teachers’ interventions. Both questions have been only partly answered (see Havranek & Cesnik, 2001), but according to Doughty, it seems that teachers’ recasts⁶ of learners’ non-native-like forms following immediately those forms are most useful in cognitive processing at the level of macroprocesses.

The above outlined SLA models make a clear opposition between the focus on meaning and the focus on form. According to them, L2 acquisition is possible when learners are first and foremost focused on meaning (see The Comprehensible Input Hypothesis and the Interaction Hypothesis), or when the focus is on meaning but the learners’ attention is simultaneously focused on relevant