Usage-Based Perspectives on Second Language Learning

Teresa Cadierno, Søren Wind Eskildsen, Teresa Cadierno, Søren Wind Eskildsen

• Publisher: De Gruyter Mouton
• Year: 2015
• Print ISBN: 9783110377323
• eBook ISBN: 9783110393255
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This edited volume brings together perspectives that find mutual kinship in a view of language as an embodied, semiotic, symbolic tool used for communicative and interactional purposes and an understanding of language use as the preeminent condition for language learning – perspectives that we conjoin under the umbrella term of usage based perspectives.

Part I: Perspectives on usage in L2 learning and teaching

Brian MacWhinney

Multidimensional SLA

Complex natural phenomena, such as human language, are shaped by processes operating on very different scales in both time and space (MacWhinney 2015). Consider the case of timescales in Geology. When geologists study rock outcrops they need to consider the results of general processes such as vulcanism, orogeny, glaciation, continental drift, erosion, sedimentation, and metamorphism. Within each of these larger processes, such as vulcanism, there are many microprocesses operating across smaller timescales. For example, once the pressure in the magma chamber reaches a certain level, there can be a slow outpouring of lava or sudden explosions. Pressure can be released through steam vents with geysers operating at regular intervals. The lava may enter lakes or oceans forming pillows or it may rest in underground chambers forming columnar basalt. The variations in these volcanic processes and their interactions with each other and plate tectonics are extensive. The same is true of human language. Within human populations, the ability to articulate and process sounds has emerged across millennia of ongoing changes in physiology and neurology. Within particular language communities, ongoing change is driven by language contact, dialect shift, and group formation. Within individuals, language learning involves a continual adaptation for both first and second languages. Within individual conversations, all of these forces come together, as people work out their mutual plans, goals, and disagreements, using language. Each of these space-time frames interacts with the others at the actual moment of language use.

To fully understand the process of second language acquisition (SLA), we must place it within this multidimensional context, both theoretically and practically. In this paper, I will attempt to show how looking at SLA through this lens forces a radical restructuring of our understandings about how languages are learned and how learning can be optimized. Most importantly, the view of language learning as emerging from a multidimensional competitive integration allows us to formulate new methods for second language pedagogy.

To explain these linkages between theory and practice, we should first consider what it means to view second language learning as a multidimensional emergent process. This emergentist approach rests on three basic principles often expressed in systems theory (Beckner et al. 2009; von Bertalanffy 1968). These are the principles of competition, hierarchical structure, and timeframes. Each of these principles plays an important role in understanding second language learning as a multidimensional emergent process, and in helping us understand how we can optimize this process.

1 Competition

Competition is fundamental to biological processes. Darwin (1859) showed how the evolution of the species emerges from the competition between organisms for survival and reproduction. The three basic evolutionary processes Darwin identified were proliferation, competition, and selection. Proliferation generates variation through mutation and sexual recombination. Organisms with different compositions compete for resources or rewards such as food, shelter, and the opportunity to reproduce. The outcome of competition is selection through which more adaptive organisms survive and less adaptive ones disappear.

The combined operation of proliferation, competition, and selection is the major engine driving change in all biological and social systems. Emergentist approaches to language (MacWhinney 1999) also view linguistic structures as arising from the processes of proliferation and competition. For the organism as a whole, the fundamental functional pressure is to reproduce. For language, the fundamental functional pressure is to communicate efficiently in ways that allow the listener to decipher messages efficiently and accurately. As MacWhinney, Bates, and Kliegl (1984: 128) noted, “the forms of natural languages are created, governed, constrained, acquired and used in the service of communicative functions.”

The handmaiden of competition is cooperation. As Bates and MacWhinney (1982) observed, humans have a great many ideas that they would love to express all at once, but language only allows us to say one thing at a time. One way in which language addresses this problem is by allowing motives to form coalitions. Bates and MacWhinney (1982) characterized the possible solutions to competition as: (1) peaceful coexistence, (2) divide-the-spoils, and (3) winner-take-all. Based on this analysis and the linkage of cue strength to cue validity (Brunswik 1956), MacWhinney, Bates, and colleagues (MacWhinney and Bates 1989) conducted a series of empirical investigations of cue processing in first and second language learning called the Competition Model, which has now resulted in over 100 published empirical studies (see http://psyling.psy.cmu.edu/papers for a bibliography). This research program has shown how second language learning emerges from competitions between first language cues and the reliability and costs of cues in the new language. Although Competition Model work has given us a good understanding of the core process of competition, it has not yet come to grips with the full multidimensional nature of second language learning.

2 Hierarchical Structure

Language is multidimensional in terms of both structure and process. The traditional levels of linguistic structure include phonology, morphology, lexicon, and syntax. Beyond these, there are language structures for conversational interaction, mental model construction, and sociolinguistic group formation. All of these structures emerge over time in response to external factors from human physiology and society. In neural terms, language use relies on virtually every region of the brain from the hippocampus, basal ganglia, and cerebellum to the anterior temporal and frontal cortex. In physical terms, much of the action is located in our vocal tract, face, tongue, and lungs, but we also use our hands, eyes, and posture to supplement the message with gesture and other signs. To decode these complex messages we rely on audition, vision, and the other senses. Social forces also shape language through processes such as migration, social class, and memesis.

Complexity arises from the hierarchical recombination of small parts into larger structures. For biological evolution, the smallest parts are the genes. For the brain, the smallest parts are the neuronal assemblies that generate competing ideas (D. T. Campbell 1960; Edelman 1987). In his seminal article entitled “The Architecture of Complexity”, Simon (1962) analyzed higher-level cognitive processes as hierarchically structured combinations of elementary information processes. These elementary pieces are configured in modules whose structure is (only) partially decomposable.

2.1 An Example

These basic architectural principles can be illustrated by the four levels of structure that emerge during protein folding (N. A. Campbell, Reece, and Mitchell 1999; MacWhinney 2010). In this process, the primary structure of the protein is determined by the sequence of amino acids in the chain of RNA used by the ribo-some as the template for protein synthesis. This sequence conveys a code shaped by evolution; but the physical shape of a specific protein is determined by processes operating after initial RNA transcription. The next structure to emerge is a secondary structure of coils and folds created by hydrogen bonding across the amino acid chain. These forces can only shape the geometry of the protein once the primary structure emerges from the ribosome and begins to contract. After these secondary structures have formed, a tertiary structure emerges from hydro-phobic reactions and disulfide bridges across the folds and coils of the secondary structures. Finally, the quaternary structure derives from the aggregation of poly-

peptide subunits based on the ternary structures. This final structure allows each protein to serve its unique role, be it oxygen transport for hemoglobin or antigen detection for antibodies. In this partially decomposable emergent system, each level involves a configuration of components from lower levels, but the biochemical constraints operative on each level are unique to that level and only operate once that level has emerged during the process of folding. If a given protein operates successfully, it promotes the adaptation of the whole organism, eventually leading to evolutionary selection for the DNA sequence from which it derives. This can be viewed as a type of backwards or downwards causality between levels (Andersen et al. 2000).

As we have seen, protein folding produces emergent structures that are then subject to the mechanisms operating at these new structural levels. The same is true for language. When articulatory gestures combine into words, they produce structures that link to meanings through emergent lexical patterns. Once words are available, they can be joined into combinations that are then subject to new mechanisms such as the tendency to place related items next to each other, as stated in Behaghel’s Law (Behaghel 1923). Different structural levels trigger the operation of different mechanisms of emergence, such as episodic encoding, generalization, topological organization, structure mapping, or common ground. For a discussion of mechanisms of emergence see MacWhinney (in press)

2.2Interlocking Linguistic Hierarchies

The principles of elementary units, partial decomposability, level-specific constraints, and backwards causality also apply to the study of language learning, where the interactions between levels and timeframes are so intense. The linguistic systems of auditory phonology, articulatory phonology, lexicon, syntax, mental models, and communicative structure are represented in partially distinct neuronal areas (Hagoort 2013) and each displays hierarchical composition between levels. For example, lexical items are composed of syllables that group into prosodic feet to produce morphemes. Morphemes combine to produce compounds, derivations, and longer formulaic strings (Sidtis 2015). Articulatory form emerges from motor commands that group hierarchically into gestures that eventually produce syllabic structures. Syntactic patterns can be coded at the most elementary level in terms of item-based patterns, which then group on the next level of abstraction into constructions, and eventually general syntactic patterns. Mental models are structured in terms of grammatical roles that link to embodied interpretations of events (MacWhinney 2008c). At the most elementary level, communicative structures involve speech acts that can group into adjacency pairs from which emerge higher-level structures such as topic chains and narrative structures. Each of these hierarchies is tightly linked to others. For example, syntax and lexicon are tightly linked on the level of the item-based pattern and in terms of the local organization of parts of speech in the lexicon (Li, Zhao, and MacWhinney 2007). Given the interactive nature of these interlocking hierarchies, full decomposition or reductionism is clearly impossible. Instead, the primary task of systems analysis is to study the ways in which the various levels and timeframes mesh. To express these interactions in the terms of the Competition Model, we need to measure the strength of competing forms or patterns and their interactions during both online and offline processing (Labov 1972).

This view of language as an emergent hierarchy has important consequences for SLA theory and practice. The basic hierarchy emerges during first language learning. However, for each level, the second language learner must acquire new structures and linkages. In part, these new structures can be acquired from the natural process of language interaction (Krashen 1982). However, as discussed in detail in MacWhinney (2012), adults face a set of risk factors that reduce the effectiveness of mere exposure to L2 input. These are the factors of entrenchment, negative transfer, parasitism, misconnection, and isolation. Entrenchment arises from the fact that linguistic structures become locked into place in cortical maps that are then resistant to restructuring during second language learning. Negative transfer arises from the inappropriate use of a first language structure in the second language. Paras...

Table of contents

• cover
• Titel
• Impressum
• Inhalt
• List of contributors
• Advancing usage-based approaches to L2 studies
• Part I: Perspectives on usage in L2 learning and teaching
• Part II: The role of frequency and exposure in L2 learning
• Part III: Development of L2 interactional and constructional competence
• Part IV: Usage-based L2 pedagogy
• Part V: Synthesis
• Subject Index
• Fußnoten