Sociolinguistic perspective on an SLA theory of mind *.

As we consider the mind that acquires a second language -- when we ask what is innate to that mind, or whether we need a specially nativist or a generally nativist theory of second language acquisition (SLA), we must remember that innateness is not the whole story. In acquisition, cognitive factors (innate or not) interact with contextual factors. Second languages (L2's) are learned over time in a process that takes place in a social context, and social factors have some impact upon that process.

In this paper I will argue that mental processes of SLA have a social dimension. Of course, we are not used to thinking of cognition as social. Most studies of the cognitive processes involved in second-language acquisition model a learner mind in the abstract, a mind in isolation: a mind alone. In this paper I will argue that an adequate theory of SLA must account for the social influences on the workings of the mind as it forms an interlanguage. In particular, I will argue that social factors affect the learner's interpretation and assimilation of second-language input, and that social factors affect a learner's production in the process of carrying on conversations. Social factors in this way affect both rate and sequence in the learner's acquisition of an L2.

1. Input to the learner is social

Lehtonen and Sajavaara (1985) made this point some time ago:

The social and cognitive organization of speaking and listening and, accordingly, the cues in the speech which give information to the interactant of the other's attitudes vary in different cultures and languages, and it is obvious that cross-cultural differences in feedback cues can lead to erroneous pragmatic interpretation of the interlocutor's intentions as a result of intercultural interference. (Lehtonen - Sajavaara 1985: 195)

Many theories of SLA lack Lehtonen and Sajavaara's essential insight, that input into cognitive processes is fundamentally social in nature. Such theories depict SLA as an essentially decontextualized process. In this view, the formation of IL is simply a logical problem for the learner. Think about the way we discuss the role of input. What matters, in this view, is the logical relationship between the syntax of the input and the syntax of the existing IL. In a way, the learner's interlanguage is viewed as a kind of partially-formed jigsaw puzzle; the learner has the logical problem of fitting new pieces, the input, into the right places in the forming puzzle. In this view, the puzzlesolver could be in any social context at all, because his/her enterprise is essentially a solitary one.

But if we think of the learner in this way as only a singular and autonomous mind, divorced from social context, then we will not see or include essential factors of social context in building our models of the learner mind. For example, we cannot adequately address in this model the essential issue of why it is that the learner attends to and incorporates some aspects of input and not others. What is it that makes certain puzzle pieces of language salient in an interaction and not others? Why is it that some pieces of input become intake and others do not? (cf. Gass's 1988 model of the relationship between input and intake) Is it solely "cognitive readiness"? Is this solely a logical problem? This paper argues that the metaphor of the L2 learner as solitary puzzlesolver has, in many ways, misled us.

An essential function of language is communication among and between people. The learner's input is provided by people in interaction, usually in a face-to-face social context. In fact, in the process of interlanguage formation, the puzzle pieces are not lying passively on a table; puzzle pieces of input come from people that the learner is interacting with. Nor is L2 input fed into the minds of passive learners the way people provide input to computers. Input is provided in the course of meaningful interactions in which the learner is both producing interlanguage and receiving target language (TL) input in a kind of dance of meaning-making. A more apt metaphor for the L2-learning mind than that of the solitary jigsaw puzzler is in fact that of the dancer. The L2 learner is learning how to move in conversations together with a partner who is willing to try to synchronize interactive moves. The dance is both constant and varied, rule-governed and context-sensitive. This dance could take place in a large group or in a dyad, but it is still a dance with other dancers whose movements are synchronized in some way with one's own. And it appears to matter a great deal to the learner who he/she is dancing with. Some people are more important to the learner than others; it is those peoples' input which is most likely to become intake, a part of the interlanguage system.

In immersion programs, children are taught all their regular curriculum through the medium of the second language. Such programs are based upon what is now turning out to be a rather simplistic assumption about learner minds and the role of input in SLA. It had been observed that children who were immersed in an L2 culture and language acquired the L2 with no apparent effort. It was assumed that if teachers provided all their input in the L2, children would be able to use their innate language acquisition abilities in the same way to painlessly and perfectly acquire the L2 in immersion classrooms. And indeed, twenty years of immersion education in Canada and the United States have produced graduates who are far superior to graduates of any other form of L2 education in terms of their fluency, pragmatic ability, pronunciation and listening comprehension. And yet over the years, teachers and observers have noted a strange phenomenon: in the upper elementary grade levels, these children who had previously made t errific progress in understanding and using the L2, seem to regress. They become resistant to speaking in French or Spanish to each other, and when they do speak their pronunciation becomes less native-like. It is suggested here that these pre-adolescents begin to find the L2 input provided by their adult teachers less useful to them, precisely because the input-providers are adults and not pre-adolescents. (And should we really be surprised that pre-adolescents want to dance with each other to a different beat than their teachers'?)

Tarone and Swain (1995) have pointed out that immersion classrooms provide children with adult input in only one (academic) register. However, children are not just autonomous language acquisition devices receiving L2 input -- they are participants in a speech community in which language varieties come to take on particular functions as the needs of the participants evolve. As children get older and become pre-adolescents they have an increased need for a vernacular variety and for language input from their peers. In immersion classrooms, they have no exposure to the L2 vernacular which they need, as pre-adolescents. The only vernacular available to them is the NL vernacular -- and this is the variety they must use if they want to switch to a vernacular. Because of this, Tarone and Swain (1995) hypothesize that immersion classrooms will become increasingly diglossic as the learners get older. In these classrooms, the L2 will come to function as a superordinate language variety, used predominantly for academic topics in conversations with teachers or for "public" discourse addressed to the class as a whole. Pre-adolescent socializing will be done in the NL vernacular. This will come about because the learners do not have the L2 vernacular input they need from the L2 interlocutors they need: peers.

Blanco-Iglesias and Broner (cf. Blanco-Iglesias -- Broner -- Tarone 1995) are conducting a longitudinal study to test Tarone and Swain's (1995) hypothesis. While their study is not yet complete, their results appear to support this prediction of increasing diglossia in the upper grade levels of immersion schools.

It appears likely that the model of second-language acquisition upon which immersion education was founded was too simple. Any input, even that consisting solely of an academic register of L2 provided by adults, was thought to be enough to trigger successful SLA by children. Pre-adolescents' social needs for input from peers in the form of a non-academic register were not considered at all. And yet immersion learners' social needs for vernacular L2 input from peers may have been sufficient to undercut the ultimate success of immersion programs in producing second-language speakers who are fully proficient in the L2, or even speakers who are willing to use the L2 when they talk with each other. There is abundant evidence that the learner's role relationships with the people who provide L2 input affects both the sort of input provided to them, and the sort of input they are willing to notice.

Of course, pre-adolescents are not the only learners for whom input is inherently social. Because input is produced by people, and people adjust their input to fit their listeners' perceived needs, it can be argued that all L2 input is a part of a social matrix. Quite a few claims have been made in recent years about the nature of input and interaction, and its importance in shaping the process of SLA. In his influential study, Long (1980) set up dyads in a room at UCLA, each with a NS conversing with a NNS. Partners had never met before and were told to just hold conversations and perform a few simple tasks. Conversations were taped. The NSs in this study were found to use foreigner talk (simpler syntax, conversational adjustments, etc.). Based on these findings, Long (Long 1983, 1985; Long -- Porter 1985) generalized that NSs (apparently all NSs in all contexts) will automatically, naturally, provide learners with modified input (i+l) in the context of interaction. This interactional modification of input w as asserted to be the force which drives their acquisition of the L2.

Long's study was an important first step, following the lead of Evelyn Hatch in moving us toward a serious approach to SLA that incorporates the construct of an interactive partner. Long's research suggested that SLA depends on interaction with an other, with a partner. But we need to remember that Long's study was conducted in a room at UCLA, perhaps the equivalent of a dance studio where every partner was apparently a willing dance partner. Outside the studio, it doesn't work that way.

Bondevik (1996), a Norwegian learner of English L2 studying in Minnesota, had read about Long's claims on foreigner talk and modified interaction and expected Americans to use foreigner talk with him when he didn't understand. But it seemed to Bondevik that in his day-to-day life in Minnesota, native speakers did not modify their input to him at all, even when he stressed that he was a foreigner and did not understand what they had just said (e.g., when he asked at a local bagel shop: "Excuse me, what are sprouts? I'm from Norway and I never heard of that", no FT occurred). To explore the factors that might or might not trigger foreigner talk, Bondevik conducted a study, not in a university room as in Long's study, but in a local electronics store. Over a period of several weeks, the same four "customers" engaged five NS electronics salesmen in conversations which were taped. The "customers" were one NS, one fluent Caucasian NNS, one less fluent Caucasian NNS, and one ungrammatical and heavily accented Asian NNS. Four of the salesmen were from Minnesota and one was from California; all were Caucasian NSs aged 25-35. At a certain point in each conversation, each customer signaled serious noncomprehension with something the salesman had said ("What's a voltage adapter?"), and the foreigners pointed out their foreignness. The salesmen's responses were all taped and analyzed to determine whether they had adjusted their speech (simplified the syntax or vocabulary, or used conversational adjustments) to accommodate to these different customers' failure to understand, as interaction theory would predict. The results? Only one native speaker (the Californian!) adjusted his speech in this way after his customers signaled noncomprehension. But the other four did not simplify either linguistically or conversationally, or use foreigner talk with the foreigners at all - even after the latter stressed their foreignness and said they did not understand. In fact, one even complexified his syntax after learners indicated noncompr ehension.

In this social context not all the native speakers used FT with the learners in the way the native speakers in the UCLA room did. Perhaps Minnesotans can't dance! Possibly the interactionists are wrong in claiming that all native speakers have the ability to simplify their speech; perhaps the Minnesota salesmen did not know how to simplify their input (after all, those of us who prepare L2 teachers suspect that the ability to adjust one's speech for learners is not innate and natural, because we know how hard it is to get beginning teachers to simplify their teacher talk for low proficiency learners). Another possibility is that the input native speakers provide is affected by different social situations; on the salesroom floor, a desire to maintain their roles as knowledgeable electronics experts in the sales encounter may have caused these salesmen to continue to use complex vocabulary and syntax with these non-comprehending customers in order to demonstrate their expertise, though the same salesmen in a un iversity experimental situation might have simplified their language with the same learners. Thus, the input provided to second-language learners is adapted but in a much more complex way than interactionists originally proposed.

Apparently we cannot assume that in social interactions outside of the lab or experimental study, NSs use foreigner talk and conversational adjustments to simplify input to L2 learners in the same way they do in experimental studies. Indeed, we know very little about the nature of the input which NSs provide to L2 learners in their midst in normal, daily interactions in authentic social contexts, or even in classrooms. Foster (1993), and Platt and Brooks (1994) suggest that the sorts of negotiations interactionists describe may not always occur much in L2 classrooms. If we are to assert that it is modified interaction and the resulting L2 input which drive the entire process of SLA, we need to know more about the nature of NS interaction and input to L2 learners in authentic social contexts.

In this section, then, it has been argued that input to the second-language learner is inherently social, that the identity and roles of the people producing that input and the context in which they produce it are important. But social factors also influence learners' production of their interlanguage as they carry on conversations with speakers of the L2.

2. It is in conversations that we learn to produce interlanguage

When producing language in any conversation, the learner tailors his/her interlanguage (IL) to fit his/her listeners: he/she style-shifts. IL varies synchronically; the L2 learner's IL may appear to be at one stage in one social context or task, and at another stage in another. The phenomenon of style-shifting in interlanguage production has been documented in many studies, some of which are described in Tarone (1988). A current model of syntactic change called Competing Grammars is being used by linguists such as Kroch (1990) and Pintzuk (1991) to account for this sort of style-shifting in language change. In this model, learners progress over time through a sequence of grammars, but at any given point in time, more than one of these grammars may be active and usable -- and a choice between these competing grammars depends on non-linguistic factors. As in dancing, the L2 learner follows a series of movements that are partially pre-determined (by the grammar in use) but are also partially locally interpreted with the partner(s). The point of learning an L2 is to learn a social skill: how to move in synchrony with others in conversation in the process of making meaning. In considering the non-linguistic causes of style-shifting, or language variation, Bell (1984) points out that cognitive factors like "attention" can only be intermediary at best. In Bell's terms, it must always be something else which causes the speaker to pay more or less attention to speech, with resulting style-shifting or variation. That something else, for Bell, is clearly a social, interactional factor: the interlocutor. The speaker shifts attention to speech, and shifts in use of language forms, depending on who the interlocutor is, the nature of the speaker's relationship to that interlocutor and the nature of the interaction. In other words, the language produced by the second-language learner is always shaped to fit the social situation, and most particularly, the relationship with the interlocutor(s). And Liu (1991) shows how those choi ces can affect the development of an IL over time.

This fact about learner language has often been minimized in our field. Again, there are many examples that could be given, but this paper will focus primarily on research on communication strategies. Researchers have struggled over the years to find a way to think about communication strategies. Early on, Tarone (1980) offered a definition of communication strategies which emphasized their inherently social, interactional nature: communication strategies were "mutual attempts of two interlocutors to agree on a meaning in situations where the requisite meaning structures do not seem to be shared" (Tarone 1980: 419). In other words, speakers, both learners and native speakers, would paraphrase, circumlocute, use nonverbal strategies and so on as part of a mutual attempt to agree on a meaning. In this definition, communication strategy use is analogous to dance: there are partners adapting to one another's moves. If we followed this definition, we would study communication strategy behavior only as it occurred in conversations between people.

But more cognitive, non-interactive definitions of communication strategies were also offered. For example, Faerch and Kasper (1983) defined them as "potentially conscious plans for solving what to an individual presents itself as a problem in reaching a particular communicative goal" (Faerch - Kasper 1983: 36). Notice that this definition has removed the social dimension; CS use is viewed as an individual problem-solving activity -- rather like our earlier jigsaw puzzle analogy. Faerch and Kasper's definition implies that it doesn't matter whether there is an interlocutor or not. Using their (and others' similar) definitions of CS, one can gather data on CS use by taping and studying isolated individuals solving verbal problems out loud in a lab.

Yule and Tarone (1997) contrast these two approaches to the study of communication strategies, terming them the Pro and Con approaches. A basic question distinguishing the approaches is, does it matter whether there is an interlocutor or not? Does communication strategy use differ when a speaker is addressing someone else as opposed to speaking into a tape recorder mike? The evidence on style shifting shows definitively that it matters. Certainly, if Bell is right, it matters. The studies in Tarone (1988) show that learners do indeed shift their IL production to fit their interlocutors, and subsequent studies such as Liu (1991) have provided even more convincing evidence that learners produce different language for different interlocutors, and that acquisition over time is affected differentially by these different interactional encounters. Definitions of communication strategies which use a puzzle-solving metaphor rather than an interactive metaphor are therefore quite simply inadequate.

As long ago as 1983, Brown and Yule presented compelling arguments for designing elicitation tasks which recognize that second-language learners are not isolated, but rather social beings whose interactional context has an impact upon their speaking performance. Brown and Yule (1983) provide guidelines for task design which stipulate that the study participant should not be asked to "generate data" in isolation, but rather to speak to an interlocutor who needs the information being conveyed for some purpose. Those recommendations have been repeated in Tarone and Yule (1989) and elsewhere (e.g., Yule 1997). Yet the notion of the decontextualized L2 learner is so strong that still today the norm is for CS studies (and other SLA studies) to use tasks for data collection in which the learner is asked to generate data into a tape recorder in a lab environment with no interlocutor present. Even experienced researchers will rationalize their choice to have their participants speak into a tape recorder (or possibly t o the experimenter him/herself even though the experimenter clearly already knows the information being conveyed, has heard it before, and has no conceivable use for it). It is clearly assumed that it does not matter who the learner is speaking to and it is just as clearly a lot of trouble to have to think about the implications for experimental design if it did matter. In other words, the learner continues to be imagined as a logical mind which generates linguistic output, totally impervious to influence from interactional context and the presence or absence of an interlocutor with a defined relationship to the speaker.

The essential point of this paper is that this is an inadequate way to think about second-language learners and the way their minds operate in the acquisition of a second language.

3. Need for an interrelated research strategy

Several strands of research need to be tied together in thinking about the mind of the L2 learner. There is:

1. research on competing grammar (Kroch 1990; Pintzuk 1991) and IL variation (Tarone 1988; Young 1991; Bayley -- Preston 1996; Mellow 1996);

2. research on neurolinguistics and SLA (Pulvermuller -- Schumann 1994; Schumann 1995)

3. research on Accommodation Theory (Giles -- Powesland 1975; Beebe -- Giles 1984);

4. research on interactional modification (Long 1985; Long -- Porter 1985; Pica 1994);

5. research on conversational co-construction (Donato 1994; Swain -- Lapkin, in press) and focus on form (Doughty 1997);

6. research on the longitudinal development of IL grammars across interactional contexts (Huebner 1983; Liu 1991).

Work on linguistic variation is related to work on interaction and input, on attention and focus on form, on co-construction in interaction, and on the development of IL forms.

In order to understand how social factors come to have an impact upon cognition and upon SLA, we need to reshape the way we understand cognition itself. It may be hard to incorporate social factors into an information-processing, computer-based conceptualization of cognition.

Schumann (1995), and Pulvermuller and Schumann (1994) point out that the brain is more than the cortex. They suggest that we should not view cognition solely as a logical information-processing system. Rather, they point to neurolinguistic work suggesting that cognition rests upon or within affect: the learner's emotional reactions to interlocutors and social and cultural contexts ("stimulus appraisal") play a critical role in cognition, and of course, in the process of SLA. They point out that many neurolinguists no longer conceptualize the mind as solely an information-processing entity, based in logic, but also as an affective and social entity, fundamentally rooted in emotional evaluation of the world about it. Such a view of cognition as emotionally and socially contextualized makes it much easier to see how social factors might interact with cognitive factors in shaping the process of SLA. (For an interesting exchange between those who treat SLA as a decontextualized process, and those who place SLA in an affective/social context, see the March 1995 volume of SSLA: Eubank -- Gregg 1995; Schumann 1995; Jacobs 1995; Pulvermuller 1995).

Giles' Accommodation Theory (cf. Giles -- Powesland 1975; Beebe -- Giles 1984) and the studies carried out within that framework certainly seem consonant with the central claim of neurolinguists such as Schumann (1995), and Pulvermuller and Schumann (1994) that at the heart of the learner's interactions with speakers of the L2 is an emotion-based approach/avoidance response. Recall that in Giles' terms, the learner has a basic tendency to either (1) identify with those interlocutors and thus converge with them in speech style, so that the learner's IL speech forms become more similar to the speech forms of the L2 speakers, or else to (2) refuse to identify with them, and thus diverge from them in terms of speech style, so that the learner's speech forms continue to be identified with older IL norms. The choice between competing grammars may in part be a choice between convergence or divergence with the interlocutor. Giles' constructs of convergence and divergence (and the studies which support them) seem comp letely compatible with Schumann's (1995), and Pulvermuller and Schumann's (1994) view of mental processes as emotion-based and grounded in stimulus-appraisal.

Accommodation Theory can provide us with a general picture of the social and interactional processes involved in SLA, but precisely how do those processes affect cognition in the L2 learner? We need detail on the dynamics involved. Hatch (1978) discussed those dynamics in terms of Scollon and Scollon's notion of scaffolding, and interactionists (e.g., Long 1980) have discussed them in terms of interactional modifications. More recently, research on focus on form has begun to examine the detail of scaffolding and recasts in communicative exchanges (Williams -- Doughty 1997). In similar work, some influential SLA researchers have turned to a Vygotskyan framework in trying to understand the mechanism by means of which specific language forms can be modeled by the interlocutor and internalized by the learner (lantolf -- Appel 1994; Platt -- Troudi 1997). Vygotsky (1986) provides a useful framework for conceptualizing the way in which convergent social interactions with valued others in differing social contexts c an shape linguistic performance and acquisition itself. Interlocutors model and support the learner's linguistic performance within that range of activities which the learner can do, but only with assistance -- this range of activities is the learner's "zone of proximal development". Stages of acquisition clearly exist; the ZPD occurs at the interface between stages, or at the point where the learner chooses between two competing grammatical structures. The result of this "co-construction process ... [is] change among and within individuals during joint activity" (Donato 1994: 39). The new forms and structures are worked out with the support of valued others in convergent interaction, and then those external activities become internalized, and part of cognition.

Swain and Lapkin (in press) describe the way in which two language learners interacted with each other in their IL to plan a subsequent oral report. The dialogue between Kim and Rick showed clearly that each had competing grammatical rules and lexical choices within their ILs. Was it 'reveille-matin' or reve-matin'? Was this set of verbs reflexive or not? Swain and Lapkin identified 32 points at which the learners helped each other choose between competing grammatical rules and items; the resulting choices always marked change in someone's IL, and usually produced more accuracy.

Of the 32 cases of collective scaffolding observed in the planning session, 75% of the language structures involved in the scaffolding were used correctly the next week. Thus, through collaborative dialogue of this sort, learners added to their own L2 knowledge and extended that of their peers. Learners provided the necessary support for each other to outperform their competence, and in the process develop their interlanguage. (Swain -- Lapkin n.d.: 9)

Finally, longitudinal studies such as Liu (1991) demonstrate how learners' participation in such interactions with interlocutors in different role relationships, over a period of time, can ultimately affect both the rate and sequence of acquisition of an L2. Liu describes a learner whose acquisition of English questions was so affected by his interactions in one social context that two of the supposedly universal stages of acquisition were altered for this learner.

4. Conclusion

In conclusion, it appears to me that there is mounting evidence that we should view the L2 learner, not within a framework which imagines the mind of the learner to be simply a decontextualized information-processing mechanism, but rather within a theoretical framework which combines the work of linguists and researchers like Kroch, Schumann, Giles, Lantolf, Doughty, Swain, Lehtonen and Sajavaara to view the mind of the L2 learner as essentially social. For such a mind, second-language input is socially mediated; L2 input is never neutral, but rather takes on social meaning for the learner. This in turn affects the acquisition process. For such a mind, IL production is also socially mediated, since production rarely takes place in isolation but occurs in conversations. The L2 learner acquires an interlanguage in the process of carrying on conversations. The learner co-constructs utterances when converging with others, in conversations which provide opportunities to hear and choose from a variety of language s tructures while making meaning. The learner's grammatical choices vary depending on his relationship with the interlocutor and the co-construction processes engaged in the conversation. The learner acquires an interlanguage in a social context as an evaluating, converging or diverging interactant, whose cognitive capacities are affected by that social context. In short, the process of L2 learning is like learning a new dance, and that process, while it is systematic and rule-governed, also has an inherently social dimension. All our thinking about the process of second-language acquisition, all our data collection procedures, all our interpretations of research results, all our model-building, should take that essential fact into account.

* An earlier version of this paper was presented at thc Second Language Research Forum, Michigan State University, East Lansing, Michigan, Oct. 19, 1997. It is offered to this volume in tribute to thc work of Prof. Kari Sajavaara, which has effortlessly bridged thc chasm between sociolinguistic and psycholinguistic perspectives on thc study of cross-language communication.

REFERENCES

Appel, G. -- J. Lantolf

1994 "Speaking as mediation: A study in L1 and L2 text recall tasks", Modem Language Journal 78: 437-452.

Bayley, Robert -- Dennis Preston (eds.)

1996 Second language acquisition and linguistic variation. Amsterdam: John Benjamins.

Beebe, Leslie -- Howard Giles

1984 "Speech accommodation theories: A discussion in terms of second language acquisition", International Journal of the Sociology of Language 46: 5-32.

Bell, Alan

1984 "Language style as audience design", Language in Society 13: 135-204.

Blanco-Iglesias, Susana -- Maggie Broner -- Elaine Tarone

1995 "Observations of language use in Spanish immersion classroom interactions", in: Lynn Eubanks -- Larry Selinker - Michael Sharwood-Smith (eds.), 241-254.

Bondevik, Stein-Gunnar

1996 Foreigner talk revisited: When does it really occur and why? [Unpublished MA thesis, University of Troms[phi], Norway.]

Brown, Gillian -- George Yule 

1983 Teaching the spoken language. Cambridge: CUP.

Cook, Guy -- Barbara Seidlhofer (eds.)

1995 Principles and practice in applied linguistics: Studies in honour of Henry Widdowson. Oxford: OUP.

Donato, Richard

1994 "Collective scaffolding in second language learning", in: Jim Lantolf -- Gabriele Appel (eds.), 33-56.

Doughty, Cathy

1997 "Meeting the criteria of focus on form", (Paper presented at the Second Language Research Forum, Michigan State University.)

Eubank, K. -- K. Gregg

1995 "'Et in amygdala ego?': UG, (S)LA, and neurobiology", Studies in Second Language Acquisition 17, 1: 35-58.

Eubanks, Lynn -- Larry Selinker -- Michael Sharwood-Smith (eds.)

1995 Festschrift for William Rutherford. Amsterdam: John Benjamins.

Faerch, Claus -- Gabriele Kasper

1983 "Plans and strategies in foreign language communication", in: Claus Faerch -- Gabriele Kasper (eds.), 20-60.

Faerch, Claus -- Gabriele Kasper (eds.)

1983 Strategies in interlanguage communication. London: Loagman.

Foster, P.

1993 "Discoursal outcomes of small group work in an EFL classroom: A look at the interaction of non-native speakers", Thames Valley University Working Papers in English Language Teaching 2:1-30.

Gass, Sue

1988 "Integrating research areas: A framework for second-language studies", Second Language Research 9, 2: 99-117.

Gass, Sue -- Carolyn Madden (eds.)

1985 Input in second language acquisition. Rowley, MA: Newbury House.

Giles, Howard -- P. F. Powesland

1975 Speech style and social evaluation. London: Academic Press.

Hatch, E.

1978 "Discourse analysis and second language acquisition", in: E. Hatch (ed.), 402-435.

Hatch, E. (ed.)

1978 Second language acquisition: A book of readings. Rowley, MA: Newbury House.

Hudson, Thom

1993 "Nothing does not equal zero: Problems with applying developmental sequence findings to assessment and pedagogy", Studies in Second Language Acquisition 15, 4: 461-494.

Huebner, Tom

1983 A longitudinal analysis of the acquisition of English. Ann Arbor, MI: Karoma.

Jacobs, Bob

1995 "Dis-integrating perspectives of language acquisition: A response to Eubank & Gregg", Studies in Second Language Acquisition 17, 1: 59-64.

Kasper, Gabriele -- Eric Kellerman (eds.)

1997 Communication strategies: Psycholinguistic and sociolinguistic perspectives. New York: Longman.

Kroch, Anthony

1990 "Reflexes of grammar in patterns of language change", Language Variation and Change 1:199-244.

Lantolf, Jim

1995 "Second language acquisition thoery?", (Paper presented at the British Association of Applied Linguistics Conference, Southampton, September 14, 1995.)

Lantolf, Jim -- Gabriele Appel (eds.)

1994 Vygotskyan approaches to Second Language research. Norwood NJ: Ablex. Lehtonen, Jaako -- Kari Sajavaara

1985 "The silent Finn", in: Deborah Tannen -- Muriel Saville-Troike (eds.), 193-201. Liu, Guo-qiang

1991 Interaction and Second Language Acquisition. [Unpublished PhD dissertation, La Trobe University, Australia.]

Long, Michael

1980 Input, interaction and second language acquisition. [Unpublished PhD dissertation, UCLA.]

1983 "Native speaker/nonnative speaker conversation and the negotiation of comprehensible input", Applied Linguistics 4, 2: 126-141.

1985 "Input and second language acquisition theory", in: Sue Gass -- Carolyn Madden (eds), 377-393.

Long, Michael -- Pat Porter

1985 "Group work, interlanguage talk, and SLA", TESOL Quarterly 19: 207-228. Markee, Numa

1994 "Toward an ethnomethodological respecification of second-language acquisition studies", in: Elaine Tarone -- Sue Gass -- Andrew Cohen (eds.), 89-117.

Mellow, Dean

1996 "A cognitive account of L2 development: The automatization of form-function mappings", (Paper presented at the Second Language Research Forum, Arizona, Oct. 24, 1996.)

Pica, Teresa

1994 "Research on negotiation: What does it reveal about second-language learning conditions, processes, and outcomes?", Language Learning 44: 493-527.

Pica, Teresa -- Cathy Doughty

1985 "Input and interaction in the communicative language classroom: A comparison of teacher-fronted and group activities", in: Sue Gass -- Carolyn Madden (eds.), 115-132.

Pintzuk, Susan

1991 Phrase structures in competition: Variation and change in Old English word order. [Unpublished Ph.D. dissertation, University of Pennsylvania.]

Platt, Elizabeth -- F. Brooks

1994 "The 'acquisition-rich' environment revisited", The Modern language Journal 78: 497-511.

Platt, Elizabeth -- Salah Troudi

1997 "Mary and her teachers: A Grebo-speaking child's place in the mainstream classroom", Modern Language Journal 81, 1: 28-49.

Preston, Dennis

1989 Sociolinguistics and second language acquisition. Oxford: Blackwell. Pulvermuller, Friedemann

1995 "What neurobiology can buy language theory: A response to Eubank & Gregg", Studies in Second Language Acquisition 17, 1: 73-78.

Pulvermuller, Friedemann - John Schumann

1994 "Neurobiological mechanisms of language acquisition", Language Learning 44: 681-734.

Schumann, John

1995 "Stimulus appraisal and second language acquisition", (Paper presented at the Second Language Research Forum, Cornell, New York.)

Swain, Merrill

1993 "The output hypothesis: Just speaking and writing aren't enough", Canadian Modern Language Review 50, 1: 158-164.

Swain, Merrill - Sharon Lapkin

in press "Interaction and second language learning: Two adolescent French immersion students working together", Modern Language Journal 82, 3.

Tannen, Deborah -- Muriel Saville-Troike (eds.)

1985 Perspectives on silence. Norwood, NJ: Ablex.

Tarone, Elaine

1980 "Communication strategies, foreigner talk and repair in interlanguage", Language Learning 30: 417-431.

1988 Variation in interlanguage. London: Edward Arnold.

1997 "Analyzing IL in natural settings: A sociolinguistic perspective on second-language acquisition", Culture & Cognition 30, 1:137-150.

Tarone, Elaine -- Sue Gass -- Andrew Cohen (eds.)

1994 Research methodology in second language acquisition. Hilldsdale NJ: Lawrence Elbaum.

Tarone, Elaine -- Guo-qiang Liu

1995 "Should interlanguage variation be accounted for in a theory of second-language acquisition?", in: Guy Cook -- Barbara Seidlhofer (eds.), 107-124.

Tarone, Elaine -- Merrill Swain

1995 "A sociolinguistic perspective on second-language use in immersion classrooms", Modern Language Journal 79, 1:166-178.

Tarone, Elaine - George Yule

1989 Focus on the language learner. Oxford: OUP.

Williams, J. -- Cathy Doughty

1997 "Focus on form in classroom interaction", (Paper presented at a colloquium at the Second Language Research Forum, 1997, Michigan State University.)

Vygotsky, Lev Semenovich

1986 Thought and language. Cambridge, MA: MIT Press.

Young, Richard

1991 Variation in interlanguage morphology New York: Peter Lang.

Yule, George

1997 Referential communication tasks. Hillsdale, NJ: Lawrence Erlbaum.

Yule, George -- Elaine Tarone

1997 "Investigating L2 reference: Pros and Cons", in: Gabriele Kasper -- Eric Kellerman (eds.), 17-30.