It is befitting to conclude our chapter on the environment by pausing to recognize voices that have warned all along that the ingredients of input, interaction, output and feedback are no magic bullet for L2 learning. Critiques offered all along by sociolinguistically and socioculturally minded researchers should help us temper our interpretations of cognitive-interactionist arguments about the benefits afforded by experience with the linguistic environment.
The limits of the linguistic environment 77
First, it is important to remember that, besides negotiations in the imminent face of communication failure, other more positive features of interaction can provide learners with potential linguistic benefits as well. Thus, Nakahama et al. (2001) showed that a conversational task engendered lower levels of negotiation than an information-gap task (a fact that is predicted by proponents of interaction, e.g. Pica et al., 1993), but that it also engendered more personal engagement and afforded many more opportunities to take risks, enabling these learners to produce longer turns and more complex language. Furthermore, Aston (1986) warned that interactional work is multifaceted and can accomplish things that have nothing to do with repairing communication problems. For instance, Foster and Ohta (2005) rightly note that not all other-repetitions are confirmation checks and that many are what they call a continuer, or a repetition that indicates rapport and has the function to encourage the speaker to say more, as in the case of (16):
(16) M1: I wasn’t so fat before I came to England V2: fat?
M3: yeah, but now I eat a lot of bread.
(Foster and Ohta, 2005, p. 421)
Second, the significance of negotiation work for the interlocutors may be deceptive. Hawkins (1985) showed that learners may feign understanding in order to avoid lengthy and cumbersome negotiations, and they may do so motivated by the very human need to be polite and save face. In her study, responses to comprehension checks (e.g. uh-huh, or an echoic repetition) were on the surface appropriate. When the two L2 participants were asked, it turned out that in fact about half of their investigated responses were issued as a polite way out. They were ‘nothing more than repeating it because he [the interlocutor] was saying it’, as one of Hawkins’ participants put it (p. 173).
Furthermore, extremely low levels of negotiation may characterize some groups and settings, as discovered by Foster (1998) in a study with 21 part-time, intermediate-level English students in a college in the United Kingdom. In that classroom, and perhaps in many others, negotiation may be almost non-existent, perhaps once again out of politeness, or maybe disengagement. Alternatively, students may be engaged in the content being learned but produce minimal negotiation, presumably because comprehension problems do not arise and inaccuracies are forgiven, as both Musumeci (1996) and Pica (2002) have documented for college-level Italian and English content-based L2 instruction, respectively. In addition, negative and even confrontational interaction patterns can occur in learner–learner work as well, as shown in the work of Neomy Storch at the University of Melbourne (e.g. Storch, 2002). Conversely, a highly motivated partner may boost the other student’s willingness to interact and engage with a given oral task (e.g. Dörnyeiand Kormos, 2000). In general, then, unfavourable attitudinal and affective predispositions can make one interlocutor or the other uninterested in negotiation or they can make negotiation moves that do occur fruitless for comprehension, much more so for acquisition.
It should be clear that not all things that can go wrong in interactions can be blamed on the non-native speaker and their level of interest, engagement, politeness, cooperativeness or willingness to learn. Ehrlich et al. (1989) found that not all people who happen to have grown up with a language are equally skilful at using it to deliver the kinds of explanations required from the tasks that researchers typically use in cognitive-interactionist studies. A particular style of L1 speaker, who they called embroiderers, engaged in confusing excess of information, causing less successful communication.
Power and prejudice are also unexamined factors in cognitive-interactionist work about the environment. Lindemann (2002) was able to show experimentally that some native speakers hold pre-existing negative attitudes against non-native- speaking groups, and that these biased attitudes can affect the processes and outcomes of concrete interactions. In her study, attitudinally biased native- speaking interlocutors approached learner utterances as problematic by default. Some of them therefore avoided negotiating at all, even when they showed willingness to negotiate actively on a similar task with another native-speaking interlocutor. Actual comprehension problems affected task outcomes (the drawing of a map route) only for those dyads where such negotiation avoidance was observed. Nevertheless, when asked, all prejudiced interlocutors evaluated task outcomes as a failure rather than a success, even when their routes had been accurately charted.
Finally, and related to issues of power, a neglected consideration is that many people who use an L2 may take what we may call an equitable responsibility approach to communication. As Schmidt (1983) astutely points out, sometimes ‘the nonnative may simply not accept the fairness of greatly disparate levels of effort by conversational partners’ (p. 167). According to Schmidt, Wes was one of them. He ‘expected native speakers to learn his interlanguage, … to consider it the native speaker’s problem as much as his own’ if mutual understanding was not achieved. More generally, and whatever language learning may come about as a result of interaction, interpersonal communication is never just about language, but it always involves interlocutors’ sense of self as well as power differentials, as we will be able to see in Chapter 10.
While all the caveats raised in this section should temper utopian or deterministic views about what the linguistic environment has to contribute to language learning, they do not invalidate the claims made by cognitive- interactionists and documented across a large body of work reviewed in this chapter. Many of these insights have also been useful in the development of research on task-based language learning, an area that has continued to burgeon to this date (Ellis, 2003; Van den Branden, 2006; Samuda and Bygate, 2008). Nevertheless, the most important contribution (and the most important limitation) that we can remember as we leave this chapter is the realization that what matters in the linguistic environment is not simply ‘what’s out there’ physically or even socially surrounding learners, but rather what learners make of it, how they process (or not) the linguistic data and how they live and experience that environment. In Chapter 5, we will examine how learner cognition exerts a powerful influence on
Summary 79
what gets processed (or not) in the linguistic data. In Chapter 10, we will explore how social structures and individual agency also shape lived experience in a dialectic of tension and, in the process, help explain the learning (or not learning) of additional languages.
4.13 SUMMARY
A few generalizations that can be remembered from this chapter are:
● The five environmental ingredients that together contribute to (but do not guarantee) optimal L2 learning are: acculturated attitudes, comprehensible input, negotiated interaction, pushed output, and a capacity, natural or cultivated, to attend to the language code, not just the message. These five ingredients were likely present in a case like Julie (see Chapter 2, section 2.2), the first of several exceptionally successful learners discovered since the mid-1990s. The last ingredient, attention to the language code, was fundamentally missing from a case like Wes (see Chapter 4, section 4.1), who epitomizes the frequently attested phenomenon of mixed learning success.
● Neither positive attitudes towards the target language and its speakers nor abundant and meaningful comprehension of L2 messages are in and of themselves sufficient for second language learning to be successful, although both are certainly important ingredients in a highly complex environmental equation.
● For successful grammar acquisition, attention to form is probably necessary. This attentional focus on form can be externally achieved by instruction or internally sought by self-study and self-directed analysis of the linguistic material available in the environment.
● Negotiation for meaning, other- and self-initiated output modification, negotiation of form during collaboration, and negative feedback of varying degrees of explicitness all carry potential for learning, provided they occur under optimal conditions that recruit attention to the language code. They facilitate psycholinguistic and metalinguistic processes of segmenting the input, noticing gaps and holes, parsing messages syntactically, monitoring and hypothesis testing; these are in turn processes that help L2 learners crack the language code.
● In the process of collaboratively negotiating choices of the code, L2 users generate language-related episodes (LREs) or pre-emptive focus-on-form episodes which are potential sites for learning. Using the L1 or metalanguage during these events may aid learning.
● Cognitive-interactionist researchers agree that negative feedback (or the implicit or explicit indication that some part of an utterance is
ungrammatical) is better overall than entirely ignoring errors. Much less agreement has been reached as to when, how and why negative feedback works, when it does.
● Recent evidence strongly suggests that the effectiveness of negative feedback is moderated by at least two factors: the degree of explicitness with which it is implemented, and the wider instructional orientation towards language as a meaning-making tool or as object for learning.
Three more subtle observations deserve some pondering:
● The value of comprehension versus production for acquisition is an ill- understood conundrum that causes disagreement among SLA researchers. Some view learning as driven by comprehension exclusively and assign production a role for fluency-building. Others claim that productive, meaningful language use is in itself a catalyst for learning.
● Grammatical competence appears to evolve in ways that are less amenable to incidental benefits from the environment than other aspects of the language to be learned, such as vocabulary, discourse competence, and so on. It also seems to hold a special status in language acquisition. Specifically, grammar (a) requires more interest, attention and hard work than other aspects of the language to be learned; (b) may even require more time to simmer and deploy than the learning of other aspects of an L2; and (c) can act as a gatekeeper to development in other areas of the L2 beyond formulaic repertoires, particularly sociolinguistic competence.
● What matters in the linguistic environment is not simply ‘what’s out there’ physically or even socially surrounding learners, but rather what learners make of it, how they process (or not) the linguistic data and how they live and experience that environment.
4.14 ANNOTATED SUGGESTIONS FOR FURTHER READING
The amount of SLA literature that has been (and continues to be) published on the linguistic environment and L2 acquisition is daunting. For this reason, a selective combination of seminal and recent studies is perhaps the best approach to this area.
You can begin by reading some of the overviews of the topic that have been written by its three most seminal proponents: a comprehensive book by Gass (1997), an oft-cited review by Long (1996) or two analytically detailed and pedagogically relevant reviews by Pica (1992, 1994). Following that, you will be able to fully appreciate the three field-influential readings by Schmidt: Wes’s case study (1983), R’s learning of Portuguese (Schmidt and Frota, 1986) and his most accessible explanation of noticing (Schmidt, 1995). Finally, your theoretical journey of the environment cannot be complete without reading the empirical study that
Annotated suggestions for further reading 81
gave rise to the Pushed Output Hypothesis (Swain, 1985). If you also then read Swain (1995) and (2000) in chronological order, you will be in a better position to understand the nuanced intellectual distance between traditional and sociocultural views of the environment (which we will explore in Chapter 10).
The empirical studies that one could read are many. The classics that marked a turning point in the field can be a good place to start: Gass and Varonis (1994), Loschky (1994) and Mackey (1999). A few recent investigations are also particularly original and felicitous (although not always easy to read). The study of LREs by Jessica Williams (1999; see also Williams, 2001) is unique in its longitudinal focus on eight students whose LREs were documented over an entire semester of instruction and 65 hours of video-taped lessons. Some of the best exemplars of recent cutting-edge research on interaction are Iwashita (2003), who examined both negative feedback and positive evidence with regard to L2 Japanese structures in a unique product-plus-process design, and McDonough (2006), who was the first to consider the potential of syntactic priming (a psycholinguistic phenomenon well studied in the L1) as a source of implicit learning through negative feedback. By contrast to the highly literate populations of college students that are featured in most research programmes, Bigelow et al. (2006) replicated an ingenious study by Philp (2003) and sought to explore the validity of cognitive- interactionist theories when they are applied to semi-literate learners. A new collection by Mackey (2007) gathers forward-looking studies of interaction across several target languages, and another by DeKeyser (2007a) covers a wide range of cognitive-interactionist topics across several contexts.
For anyone interested in negative feedback, specifically, Chaudron (1977) and the four studies summarized in Table 4.1 are excellent starter reading. Sheen (2004) and Lyster and Mori (2006) provide interesting insights about the role of the context. Recent publications by Rod Ellis and his present and former colleagues at the University of Auckland provide a fruitful direction for future research in this area (e.g. Ellis et al., 2001, 2006). Finally, two articles by Ferris (2004) and Hyland and Hyland (2006) offer excellent overviews if you are interested in L2 writing and negative feedback, and Heift (2004) and Lai and Zhao (2006) provide good examples of technology-delivered feedback.
Cognition
Cognition refers to how information is processed and learned by the human mind (the term comes from the Latin verb cognoscere, ‘to get to know’). SLA researchers interested in cognition study what it takes to ‘get to know’ an additional language well enough to use it fluently in comprehension and production. We are far from a satisfactory understanding of second language as a form of cognition, however. This is because our capacities to investigate the relevant questions are shaped by the pace at which new theories and methods to inspect the workings of human minds and brains become available (typically in neighbouring disciplines) and the rate at which SLA researchers become conversant in them. In this chapter, more than in any other, I will make frequent reference to relevant L1 research and point at areas where future attention by SLA researchers will be needed.
It is also important to realize that, in cognitive research, the relevant behavioural and neurobiological evidence falls in the order of a few hundred milliseconds to a few seconds, or it consists of larger-scope performance that nevertheless lasts a few minutes to a few hours at the most. This is in sharp contrast with many of the data on language learning SLA researchers normally consider, which involve stretches of discourse, multi-turn interactions with human interlocutors, extended texts, referential and social meaning, and even years of studying, using or living with an L2. Thus, the differences in grain size, temporal and ontological, of the various phenomena that are brought together into cognitive explorations of L2 learning are puzzling.
In this chapter you will learn about cognitive SLA theories and constructs that have been developed to explain the nature of second language as a form of cognition. The theories can be broadly classified into traditional information processing, which has dominated SLA theorizing and research since the mid-1980s, and emergentism, which is a development of the late 1990s that grew out from the former. A central preoccupation in SLA research on cognition is with memory and attention in L2 learning.