|
Language Learning Roundtable*
7 September 2011
Theme
Is there a future for the native speaker in SLA research?
Speakers
Kenneth Hyltenstam, Stockholm University:
"The native speaker as a methodological instrument - are there alternatives?"
Mike Sharwood Smith, Heriot-Watt University, Edinburgh University, The Academy of Management, Warsaw:
"Is the native speaker a psycholinguistic impossibility?"
Antonella Sorace,The University of Edinburgh, University of Tromsø:
"The bilingual native speaker"
Discussants
Michael H. Long
University of Maryland
Silvina Montrul
University of Illinois at Urbana-Champaign
Chairs
Niclas Abrahamsson and Manne Bylund
Stockholm University
* The event will take place through a grant by the journal Language Learning.
Abstracts
Kenneth Hyltenstam, Stockholm University:
"The native speaker as a methodological instrument - are there alternatives?"
The talk will take as a starting point that the notion of “native speaker”, in spite of severe and in many cases justified challenges over the last couple of decades, is central and indispensible in much (SLA) research. The meaning of notions such as “second language” or “second language speaker”, irrespective of whether studied in early, intermediate, advanced or near-native developmental phases, or, indeed, SLA as a discipline in much of its current execution, are dependent on the relationship to or comparison with native language norms or native speaker norms.
Leaving the sociolinguistic issue of native speaker identity and related value judgments aside – not because that issue is less important to deal with seriously in research and in society at large, but because it is beyond the scope of this presentation – there are two different notions of native speaker that play a salient role in psycholinguistic second language research methodology. The first of these two notions is abstract and has nothing to do with real speakers in observable time and space. In Chomsky’s well-known formulation, this native speaker is “an ideal speaker-hearer, in a completely homogeneous speech community, who knows its language perfectly…” etc. (Chomsky, 1965:3), the idea of monolingualism as a defining characteristics being congruent with the formulation “homogeneous speech community”. This theoretical construct is relevant for the analysis of language as a formal system beyond the disparities of practical language use, and is basic not only to generative linguistics in all its conceptions, but to all structuralist approaches to the study of language from Saussure onwards, and also in varying degrees to historic linguistic paradigms before the 20th century. The structure of a certain language is defined by the native speaker’s perception/intuition and is thus neutral to such distinctions as language, dialect, sociolect etc., including inherent structural variation within any variety. It clearly has nothing to do with the sociolinguistic notion of standard language, a common confusion. Any type of general reference or comparison to native speaker behavior in second language research is rooted in this abstract notion of native speaker. As such it is widely used as a methodological tool, often implicitly, in SLA research.
The second notion concerns a real native speaker, i.e. an observable language user. In second language research this notion is relevant especially in relation to the selection of native controls. Although in everyday language people tend to understand what is meant by a native speaker, it has not been possible to arrive at a sustainable scientific definition. The issue of nativeness is complicated by a number of well-known factors, for example: Is exposure from birth necessary or not? Is the linguistic behavior of monolingual speakers different from that of bilingual speakers of the same language? Do originally monolingual speakers change their linguistic behavior in their first language after they have started to learn a foreign/second language? Can a speaker cease to be a native speaker after decreased contact with other speakers of his/her first or native language? Considering complexities of this kind, researchers have opted for different solutions. In our own research (Abrahamsson & Hyltenstam, 2009), for example, we have chosen to operationalize the native speaker of Swedish using several basic criteria (Swedish being the birth language, the language for primary socialization, the only language used as means of instruction at school, the majority language in the wider society). Others, at a more general level, similarly have attempted definitions based on a characterization of the most central members of the category native speaker within a prototype theory approach (Turner, 1997; Escudero & Sharwood Smith, 2001).
In this paper I will note that both the abstract and the real notion of native speaker, with their inherent vagueness, continue to be indispensible as research tools in many of the issues that are investigated in second language research. However, it will also be suggested that the current challenges to the notion of the native speaker, and new ongoing research on this notion is giving rise to different novel understandings and methodologies. The next step may involve a phenomenological refinement of new categories that would be based in empirical research on the outcomes of different conditions for language acquisition, and could provide alternatives to the categories native and non-native speakers.
Mike Sharwood Smith, Heriot-Watt University, Edinburgh University, The Academy of Management, Warsaw:
"Is the native speaker a psycholinguistic impossibility?"
The notion of native speaker is relatively easy to anchor in a sociolinguistic or socio-psychological sense as an identity that people with a defined background attribute to themselves and to certain other people. In strictly psycholinguistic terms, that is without having recourse to external social circumstances, it seems to make little sense when more than a vague definition is called for. This has to be an embarrassment for those who wish to use the native speaker standard as a reliable criterion against which to judge various stages of L2 development. We can indeed compare 2 speakers who have, on non-psycholinguistic grounds, already been separated out into a category of ‘native’ and another category ‘other’, e.g. late-learned L2 user, and we can come up with psycholinguistic differences and similarities. Perhaps, there is some sense in trying not to define the term too precisely. However, this option is of course open to criticism on the grounds of being unscientific. Sharwood Smith (1992) attempted to save the notion by applying various tests to it in order to distinguish very advanced, non-native speakers with those who, everyone would agree, were fully native, including slightly attrited ones, and signally failed; he accordingly ended up by declaring the native speaker, as a psycholinguistic concept, ‘dead’. In addition, a different notion was proposed, one which still required reference to a group predefined in terms of their background but which allowed a different perspective on the imprecise notion of ‘near-native. This term was ‘virtual’ native: a virtual native speaker’s performance in a wide variety of situations is consistently perceived to be fully native. This leaves open the possibility of difference that emerge only as a result of sophisticated linguistic and psycholinguistic tests that are not dependent on other people’s perceptions.
Escudero and Sharwood Smith (2001) attempted to resurrect the native speaker as a psycholinguistically respectable notion by refining it using prototype theory. Ultimately we have to face the possibility that the interaction between any language learner and their external environment may depend more upon essentially the same mechanisms that are processing qualitatively different input. The relevant place to look is at research that explores possible differences between very young simultaneous bilinguals and older sequential ones: here, in principle, the crucial input for working out the basic system should be the same. But a new problem arises: since the internal cognitive environment has changed the younger bilingual has not got a fully fledged language system in place already. This throws the spotlight back on the nature of the internal mechanisms that process language and for that we need more explicit accounts or how development actually occurs. The discussion will end with a theoretical account that tries to address just this kind of problem from a processing perspective.
Antonella Sorace,The University of Edinburgh, University of Tromsø:
"The bilingual native speaker"
The traditional point of reference in studies on adult second language (L2) acquisition has been the monolingual native speaker. In recent research, however, the scope of research has broadened from the exclusive use of first language (L1) monolingual users as the reference point for L2 speakers to a comparison between adult L2 speakers and other bilingual groups (i.e. child bilingual L1 acquirers, early and late consecutive child L2 learners, L1 speakers undergoing attrition). The interdisciplinarity of research on bilingual language development and the increasing attention to non-linguistic aspects of language competence requires the integration of theoretical insights and methodological tools from different fields - not just linguistics but also neuroscience, cognitive psychology, evolutionary biology.
This ‘big tent’ approach to the study of bilingualism puts the concept of the native speaker in a new light. While I maintain an operational definition of ‘native speaker’ as someone exposed to a language since infancy, I will discuss the native speaker from the following perspectives:
(a) L1 attrition. The L1 of individual speakers undergoing attrition from exposure to an L2 selectively changes (Tsimpli, Sorace, Heycock and Filiaci 2004; Sorace 2011): how do these speakers compare to monolinguals? Are attrited speakers ‘ex-native speakers’?
(b) the “bilingual paradox” (Bialystok 2009). Early bilingualism brings cognitive advantages in terms of mental flexibility and executive control, but bilinguals on average tend to have smaller vocabularies and slower lexical access in each of their languages than monolinguals, and also less efficient integration of syntactic and contextual information (Sorace and Serratrice 2009). How does this affect the definition of early bilinguals as native speakers of two languages?
(c) child L2 learners: While exposure to an L2 in childhood has traditionally been seen as resulting in native competence, recent comparisons of early and late successive bilingualism within childhood indicate that the acquisition of some aspects of language in terms of grammatical development and attainment depends on whether exposure to the other language happens before or after age 4 (Meisel 2009; Unsworth et al 2011). Are there two different kinds of native speakers, or only one?
(d) L2 near-nativeness. If we compare native monolinguals and very advanced L2 speakers (improperly called ‘near-native’), we see that they both show more variation with respect to linguistic structures that require the coordination of multiple types of information (Sorace 2011), although to different degrees. What are the implications for the traditional view of native speakers as a homogenous community?
|