Frequency Effects In Instructed Second Language Acquisition

Karin Madlener

• Publisher: De Gruyter Mouton
• Year: 2015
• eBook ISBN: 9783110423594
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

Based on a state-of-the-art review of prior research in all related domains, this book makes precise predictions about the expected effects of specific type and token frequency distributions in input floods and tests these in the second language classroom context.

Part I:Do all roads lead to Rome? Ways to construct a second language

1 Cognitive Linguistics meets the second language classroom

In spite of their divergent goals and research agendas, second language (L2) acquisition research and L2 teaching share a number of concerns: What can and what do learners make of specific learning conditions? Why do L2 learners often fall short of advanced, let alone full, competence? How do social, cognitive, and affective learner characteristics such as age, prior language experience, memory, and motivation interact with such aspects of the learning situation as the input, communicative load, task complexity, and degrees of interactivity and explicitness, in determining L2 learning rates, routes, and endstates?

In other words, among the many roads that could possibly lead to Rome, what makes some more reliably highway-like and others dead ends? What do specific learner populations, for example, instructed adult L2 learners as opposed to early bilingual heritage language speakers, need to successfully (re-) construct the L2 patterns? Speaking in terms of instructional settings, under which conditions will most learners in most classrooms be most likely to reliably develop good L2 competence on the long term?

This book explores the question of optimal (L2) learning conditions from a usage-based cognitive linguistic perspective. Cognitive Linguistics, of course, is not a fixed theory, but an emergent family of related approaches, ranging from Cognitive Grammar (Langacker 1987), Conceptual Metaphor Theory (Lakoff and Johnson 1980), and Cognitive Semantics (Talmy 2000) to Construction Grammar (Croft 2001; Goldberg 2006, 2013; Boas 2013), usage-based models (Bybee 2013; Diessel 2013; Ellis 2013; Tomasello 2003), and emergentist and Complex/Dynamic Systems approaches (Ellis and Larsen-Freeman 2009; De Bot, Lowie, and Verspoor 2007).

The main tenets of cognitive linguistic approaches may be summarized as follows (cf. Bielak 2011; Boers and Lindstromberg 2008; Geeraerts 2008): Firstly, language is part of general cognition, not separated from it. Secondly, language knowledge is organized into a structured network of conventionalized form- meaning mappings, so-called constructions, not a generative algorithm. Thirdly, language is an instrument for organizing, processing, and sharing information, which puts the main focus of linguistics on meaning (rather than on formal paradigms) and on multiword units (rather than on single words). In general, CL approaches do not assume strict separations either between lexicon and grammar or between pragmatics and semantics, but a dynamic, usage- based model of language processing and acquisition.

To date, most cognitive linguistic approaches have mainly focused on linguistic description, first language (L1) processing, acquisition, and modeling, but implications and applications for L2 processing and acquisition have begun to emerge (cf. Achard and Niemeier 2004; Boers and Lindstromberg 2008; Dabrowska 2009; De Knop and De Rycker 2008; De Knop, Boers, and De Rycker 2010; Ellis 2013; Littlemore 2009; Pütz, Niemeier, and Dirven 2001; Robinson and Ellis 2008; Tyler 2012). So how can Cognitive Linguistics possibly contribute — theoretically, empirically, and methodologically — to the development of L2 acquisition research and of instructional methods and materials? More precisely, what may an applied usage-based cognitive linguistic approach contribute to the theory of L2 learning and and to the practice of L2 teaching?

1.1 Think it over: Explicit learning in the classroom

Cognitive linguistic applications to L2 acquisition have mainly been based on Cognitive Grammar, Cognitive Semantics, and Conceptual Metaphor Theory so far. They have focused on improving explicit learning and teaching, that is, pedagogical grammars, aiming at a reduction of the perceived arbitrariness of the L2 system (Taylor 2008: 57). As such, they promote enhanced noticing, analysis, awareness, and understanding of the L2 structures and conceptualizations as prerequisites for deep processing and better retention (cf. Bielak 2011: 248; Boers and Lindstromberg 2008; Holme 2010; Verspoor and Boers 2013).

This type of cognitive linguistic applications targets “insightful learning”of the underlying semantic and conceptual motivations of specific form-meaning mappings on the one hand and of the relations between constructions in the constructional network on the other hand (Verspoor and Boers 2013: 701ff). Enhanced understanding may be promoted through “descriptively adequate, intuitively acceptable, and easily accessible” explanations (Matsumoto 2008: 150) and visualizations (Scheller 2008; Suñer Muñoz 2013). A variety of applications have been proposed, for example, for polysemous prepositions and phrasal verbs (e.g., Alejo González 2010; Cho 2010; Dirven 2001; Tyler 2012), tense and aspect (Tyler and Evans 2001), and modal verbs (e.g., Tyler 2012). Central concepts used here are image schemas, figure, and ground as well as prototypes, radial networks, and family resemblances (cf. Bielak 2011). The notion of embodiment is probably a further promising option (cf. Littlemore 2009). For applications to metaphor and metonymy, see, for example, Barcelona (2001), Littlemore (2009), and Roche (2013). For applications to the learning and teaching of vocabulary and idioms, see, for example, Boers (2013) and Kovecses (2001).

In addition, cognitive linguistic applications based on Slobin’s (1996) Thinking for Speaking Hypothesis and on Talmy’s (2000) Cognitive Semantics have aimed at fostering L2 learners’ appreciation for language-specific conventionalized preferences in terms of construal and perspective taking, and for cross-linguistic differences in terms of conceptualization (cf. Bielak 2011; Holme 2009; Littlemore and Juchem-Grundmann 2010: 2). Applications have mainly been proposed in the domains of tense and aspect (e.g., Niemeier 2013; Niemeier and Reif 2008) and of localizations, motion, and positioning events (e.g., Cadierno 2008; De Knop and Dirven 2008; Gullberg 2009, 2011).

In sum, this line of cognitive linguistic applications aims at promoting the learners’ insights into the L2 system by raising their consciousness and recruiting their (metalinguistic) awareness for the conceptualizations that are conventionally (and systematically) associated with the L2 form-meaning mappings (cf. Taylor 2008: 57). As such, this mainly targets highly motivated and analytically skilled, possibly academic learners.

In terms of actual instructional implementations, this strand of cognitive linguistic applications most likely calls for a traditional PPP approach (presentation, practice, production, inspired by Skill Acquisition Theory, DeKeyser and Criado Sánchez 2012). It probably can also be implemented in a kind of Processing Instruction approach (cf. Llopis-García 2010, VanPatten 2004b) or in an explicit, inductive or deductive Focus on Form approach (e.g., using consciousness raising tasks; cf. Wong 2005). By contrast, the approach will be rather difficult to embed into meaning-focused communicative or task-based approaches (but see Achard 2008).

1.2 Between the lines: Implicit learning in the classroom

A complementary (and somewhat hesitantly emerging) strand of cognitive linguistic applications to L2 acquisition follows a usage-based, constructionist, or emergentist approach (cf. Ellis 2008; Ellis and Ferreira-Junior 2009a, 2009b; McDonough and Kim 2009; McDonough and Nekrasova-Becker 2014; Verspoor and Behrens 2011). Instead of (exclusively) focusing on explicit learning and knowledge about language, it targets the enhancement of incidental learning from structured input in untutored L2 learning and under implicit teaching conditions.

Now, we have known for ages that the concept of input is highly important for language acquisition, that “no individual can learn a second language without input of some sort” (Gass 1997: 1). Usage-based approaches assume an even more crucial role for the input, though, proposing that actually all of “our knowledge of language is ‘constructed’ on the basis of the input” (Goldberg 2009: 93—94). In this view, the major part of language learning, and also of L2 learning, takes place implicitly and incidentally during meaning-focused input processing. Acquisition, in this perspective, is understood as the piecemeal learning of constructional exemplars, the gradual entrenchment of form- meaning mappings, the frequency-biased abstraction of regularities across exemplars, and the progressive generalization of these constructional patterns (Ellis 2002: 143).

Usage-based models have successfully shown that for Lis, implicit learning from the input — via processes such as entrenchment, distributional tallying of form-function mappings, schematization, and categorization — is crucial for linguistic processing, storage, and acquisition (Behrens 2009b; Goldberg 2009; Tomasello 2003). This predicts significant effects for input features such as frequency distributions, salience, and recency on all aspects of processing, storage, and acquisition (Bybee 2006; Diessel 2007; Ellis 2002). The development of productivity in L1 acquisition and in artificial language learning (ALL) seems to be specifically dependent on input features such as type variability and type-token ratios (Goldberg 2006; Goldberg and Casenhiser 2008; Suttle and Goldberg 2011). The question whether the development of productivity in instructed L2 acquisition is bound to the same mechanisms and frequency effects is only beginning to be investigated empirically (Ellis 2011a; Ellis and Ferreira-Junior 2009a, 2009b; McDonough and Kim 2009; Year and Gordon 2009). Continuing research promises new perspectives for L2 instruction, namely for improved input structuring as a prerequisite for optimal input processing (Ellis 2009).

As a matter of fact, L2 learners most often “have to enter into communication from experience of a very limited number of tokens” (Ellis 2011a: 203), that is, of witnessed exemplars of a pattern, in comparison to most Li learners. As their input sample is generally reduced, “incomplete, uncertain and noisy” (ibid.), estimating how linguistic constructions work in a second language is all the more difficult. Providing learners with good, rich input is therefore a major task in language teaching in general, but is all the important more from a usage- based point of view. This point has actually been made by several researchers, for example, Bielak (2011), Holme (2009), and Verspoor and Boers (2013), but it has, to the best of my knowledge, not yet led to substantial empirical research and classroom applications.

A focus on the input, on contextualized exposure, and incidental learning does not deny roles for (pushed) output, for interactional scaffolding, modifications, elaboration, and alignment, for the negotiation of meaning and form, and for (corrective) feedback. But it suggests that explicit learning and teaching may have been overemphasized for adult L2 learning (Ortega 2013: 13; cf. Roehr 2010). The observation that adult L2 learners usually rely heavily on explicit learning processes (cf. Mitchell, Myles, and Marsden 2013: 98, 136) is certainly correct, but it is not necessarily due to the fact that, roughly speaking, adults have lost their capacity for implicit learning and must rely on explicit learning. It may well be that they have been trained to do so and do not get much of a chance to use implicit learning processes if materials and teaching approaches do not provide the corresponding opportunities. Just as one may conclude that explicit learning processes must be optimized (see Section 1.1), one may conclude that in addition to enhancing understanding and analysis, L2 classrooms should increase and improve occasions for incidental, implicit learning (from the input).

Of course, exposure to natural input, even to large amounts of natural input, as provided in immersion settings or in content and language integrated learning (CLIL), is often not sufficient to trigger successful acquisition — because many constructions are rare, complex, or lack perceptual salience (Ellis 2008) and because L1 processing routines are deeply entrenched and learned attention may obscure the relevant L2 cues and construals (Ellis and Sagarra 2010). A usage-based approach to L2 acquisition contributes (i) an approximation to the question of why so much of the input usually fails to become intake and (ii) a number of theoretically founded and empirically testable assumptions as to what good, rich input (which will most probably not fail to become intake) should look like and why (cf. Bybee 2008; Ellis 2009; Goldberg and Casenhiser 2008).

A usage-based cognitive linguistic approach along these lines recruits domain-general implicit cognitive processes, such as entrenchment and schematization, for language processing and learning. And most importantly, it recruits (language) cognition for communication; thus it is situated at the intersection of cognitive la...

Table of contents

• Cover
• Title
• Copyright
• Acknowledgements
• Contents
• List of abbreviations
• List of tables
• List of figures
• Part I Do all roads lead to Rome? Ways to construct a second language
• Part II A usage-based approach to instructed second language acquisition
• Part III Frequency effects in incidental learning from structured input floods: A classroom study
• Part IV Summary and discussion
• References
• Subject index
• Appendix A
• Endnote