Russell S. Tomlin

Attention in Cognitive Science and Second Language Acquisition

This paper examines how the cognitive notion of attention has been employed in SLA and how it is understood in cognitive science. It summarizes recent research on attention from cognitive and neuroscience approaches. Some reformulations of problems raised in SLA research related to attention are proposed. Current research offers detailed ideas about attention and its component processes. These ideas, elaborated theoretically and empirically in cognitive neuroscience, may help untangle some important but difficult issues in SLA. Early, coarse-grained conceptions of attention, such as the limited-capacity metaphor or the automatic versus controlled processing dichotomy, are recast into an integrated human attention system with three separate yet interrelated networks: alertness, orientation, and detection. This finer grained analysis of attention is employed in a model of the role of attention in SLA.

Functionalism in Second Language Acquisition

This article examines the role played by functional approaches to linguistics in understanding second language acquisition (SLA). Central premises and tenets of functional approaches are described, and several key theoretical problems with functional efforts are detailed. The problem of referential management (the selection of nominal vs. pronominal NPs) in second language discourse production is examined. The general conclusions are drawn that (a) functional approaches to linguistics have a significant role to play in SLA studies, but (b) functional universals are insufficiently grounded theoretically and empirically at this point to contribute more than heuristic guidance to SLA theory.

On the interaction of syntactic subject, thematic information, and agent in english

Abstract A detailed analysis of texts of the play-by-play action of ice hockey taken from videotaped games shows that for English the function of a syntactically defined subject is to encode primarily thematic information and secondarily agent. That is, when several NPs compete for the subject relation, and it can be determined that one NP is the most thematic, then that NP will occur in the subject relation independent of its semantic role. But when no NP is more thematic, the subject will then take the agent.

The Treatment of Foreground-Background Information in the On-Line Descriptive Discourse of Second Language Learners.

This paper compares the foregrounding strategies of native speakers of English and advanced learners of ESL. Fifteen native speakers and thirty-five advanced learners produced on-line (play-by-play) descriptions of the unfolding action in an animated videotape. A methodology for the quantitative analysis of discourse production is described which permits the explicit identification of foreground-background information and its interaction with both native speaker and interlanguage syntaxes. Results show that: (1) Native speakers and learners exhibit one universal discourse strategy in retreating to a more pragmatic mode of communication under the severe communicative stress of on-line description, as revealed through the systematic loss of the coding relations ordinarily holding between foreground-background information and clause independence/tense-aspect; (2) Learners also exhibit a learner-specific general communication strategy in which significant events are described but nonsignificant events avoided.

The identification of foreground‐background information in on‐line oral descriptive discourse∗

Abstract The identification of discourse relations, like foreground‐background information, in natural text data represents a critical methodological issue in research into the functions of syntax in discourse. Currently, methods of identification are either syntax‐dependent or exclusively introspective, resulting in circular or vague argumentation concerning the role of syntactic form in coding discourse relations. An empirically‐grounded methodology for identifying foreground‐background information in natural discourse data is presented and motivated. An experiment is described in which 15 native English speaking subjects described on‐line a 108 second animated videotape. Results show that independently determined event significance correlates significantly with the frequences with which subjects report events. These results permit the formulation of a syntax‐independent and operationally explicit methodology for identifying foreground‐background information in on‐line descriptive discourse.

The frequency of basic constituent orders

Abstract This paper describes the production of an empirically valid and statistically sound sample of the worlds languages, which is used to determine the frequency of each of the six mathematically possible orders of the basic constituents S, V, and O. Using the Kolmogrov test as a test of goodness‐of‐fit, a sample of 402 languages is created which shows genetic and areal similarity to a standard produced using the Voegelin and Voegelin (1977) reference work. Unlike the results of previous convenience samples, a frequency analysis of this sample reveals the following relative frequencies of basic order types: SOV = SVO VSO VOS = OVS OSV. It is argued that greater care is needed in the production of language samples for typological studies.

Cognitive foundations of second language acquisition

The growth of cognitive science as a field, or at least as a descriptive term, indicates an emerging scientific consensus that many fundamental human capacities require collaborative and interdisciplinary research to make further fundamental headway. Language, as central to our essential humanity as anything is, represents one area in which massive amounts of interdisciplinary research is underway at virtually every research institution in the world. It strikes us, then, how comparatively little interdisciplinary research there is within cognitive science about second language acquisition (SLA). There is, of course, a great deal of research within SLA itself that draws upon research in cognition and that extends those ideas in important and interesting ways into both SLA and second language instruction (SLI). Yet SLA has never really taken a particularly prominent place within cognitive science overall. Even a brief glance at recent proceedings of the Cognitive Science Society reveals comparatively little effort directed at SLA. The irony is doubled when one considers that SLA has as a discipline properly taken pride in its multidisciplinary roots. It combines ideas and research strategies from linguistics, sociology, anthropology, education, psychology, and even biology as it examines the myriad difficulties of describing and explaining how individuals fail and succeed in learning additional languages. In the winter of 1992, a symposium was held at the University of Oregon in Eugene on the topic of cognition, SLA, and SLI. The symposium was sponsored principally by a grant from the Keck Foundation, which supported a general program in cognition and instruction in the University of Oregon Institute for Cognitive and Decision Sciences.1 The purpose of this meeting was to assemble researchers in the three related areas of cognition, SLA, and SLI to try to identify and define research questions that cut across the three areas. Approximately 30 participants from around the world and representing varied theoretical approaches to SLA worked together for 3 days looking for points of contact among their various perspectives on SLA, SLI, and cognition.2 The present issue is one product of that event. The contributions in this issue span a continuum from those whose bases are principally within SLA but tied to issues of cognition to those that are largely grounded in the more general cognitive literature but tied to issues of importance in SLA. Andersen and Shirai examine the relationship between child language acquisition and SLA of tense–aspect distinctions. They argue that there are several fundamental cognitive principles underlying the pattern of acquisition seen in the data: the Relevance Principle, the One to One Principle, and the Principle of Congruence. They argue further that these principles are not peculiar to acquisition per se but are in fact principles that govern ordinary native speaker discourse interaction. Bialystok presents a distillation of key ideas from her long-standing research program on analysis and control in SLA. Both analysis and control are fundamentally cognitive ideas about the language learning capacity, and Bialystok considers how these ideas address five central problems of SLA: the similarity between first and second language acquisition, the starting point for SLA, consciousness, variability, and instruction. Birdsong introduces a new research arena for SLA. A number of years ago, Tversky and Kahnemann (1974) examined how humans routinely misjudge the frequency of events they encounter. These misjudgments stem from the employment of general heuristics for estimating how frequently events occur. Although commonly successful, these heuristics nonetheless routinely lead to mistaken and biased judgments; for example, estimates of the product of 6 × 5 × 4 × 3 × 2 tend to be higher than those of the mathematically identical expression 2 × 3 × 4 × 5 × 6 because of the anchoring point at which the estimates begin. If such biases operate in many other domains where hypotheses about specific phenomena are formed, then it is intriguing to consider how those same heuristics and biases may affect formation of interlanguage hypotheses. Tomlin and Villa present a detailed discussion of attention and SLA. While it has long been recognized that attention plays a central role in acquisition, our views of attention have often been somewhat naive, combining folk ideas of attention with more technical ones. The authors provide a detailed review of the attention literature in cognitive psychology and argue that a more detailed model of the attentional system may help SLA formulate a finer grained model of the role attention plays in the online and real-time process of acquisition. Carr and Curran present a detailed examination of implicit learning. A problem of considerable interest in SLA has been whether SLA can occur without conscious or aware engagement of input by the learner. The issue is particularly difficult to examine because of theoretical difficulties in specifying what implicit or nonconscious learning means and because of tremendous methodological problems in investigating such issues. This article reviews both the theoretical issues of defining clearly what implicit learning is and the methodological issues of investigating implicit learning carefully in limited domains. Schumann examines the neurobiological basis of the so-called affective filter in SLA. In doing so, he raises two issues of importance to SLA. First, he challenges whether research into affective constraints on SLA should take the secondary place in the field it has in recent years. Second, he challenges simplistic views of cognition, as well, by arguing that cognitive theories must recognize their ultimate connection to the architecture in which they are situated.3 In a sense, he argues from neurobiological grounds that we cannot afford to treat affective issues as secondary to rational ones in formulating SLA theory. Despite what I hope readers will find worthwhile papers, this issue is missing work that our ideal issue would have included. There are no articles that address directly issues relating cognition and SLA to instruction. The Editor and the Editorial Board of SSLA along with the volume editors believed it to be important to include research that identified issues in instruction that might be addressed by interaction of cognitive science and SLA. Such issues did form part of the symposium, and we did attempt to include this coverage in the present effort. However, circumstances in the end did not permit us to include such a paper and this remains a gap in what we have offered in this thematic issue. Perhaps another will find the bridge between cognition and instruction easier to construct, setting a new theme for a future SSLA special issue. One final observation is in order. At first glance the contribution this issue makes will once again seem mostly to tell us in SLA that we would benefit from a more careful and detailed extension of ideas and methods from cognitive science into SLA. Carr and Curran present elaborate argumentation on issues of implicit learning. Tomlin and Villa argue for an increasingly complex model of the role of attention in SLA. Birdsong links SLA to cognitive biases tied to decision making in general. Schumann examines the neurological basis for the affective filter. Such extensions may well benefit our general research mission, and we hope very much that readers will find these contributions useful. But these same papers also make a case for incorporating the study of SLA more directly into the research agenda of cognitive science. SLA is a process in which matters about language that are ordinarily confounded may be separated. Child language acquisition faces unavoidable confounding of language acquisition with other aspects of child cognitive and biological development. The variability inherent in interlanguage grammars poses a real challenge to tidy theoretical conceptions of the nature of linguistic representations. The complication of examining the interplay between SLA and SLI may help clarify enduring problems about the interplay between social interaction (teaching) and cognitive change (learning). In these areas, along with many others, SLA has as much to offer cognitive science as cognitive science does SLA. Put more simply, SLA has always been a part of cognitive science, and we, researchers in this vital area, very much cognitive scientists.