The Role of Psychology in Applied Linguistics
The role that the psychology of language plays in applied linguistics provides a good example of the strengths and weaknesses of theoretical, scientific inputs into the process of the design of language teaching. There are several ways of determining the role and scope of the psychology of language in this regard. One way is simply to look at what is actually being done in the field, and to try to classify this, in a more or less systematic manner, as best one can. If this is one’s method, a vast array of studies on various aspects of syntactic, morphological and conversational skill acquisition in both first and second language immediately presents itself (for a survey of earlier studies, cf. Cook 1978; Clark and Clark 1977). In the case of second l anguage learning research, there is also, of course, the investigation of language learning, of the perception and production strategies and of motivational and situational (mostly pedagogical) variables. But the overwhelming magnitude of research studies done would soon make it clear that confusion would reign supreme if one tried to define the scope of a psychology of language only by looking at what is being done, without also asking what it is that stimulates this work, i.e. consider the question why it is being done (Paltridge 2014).
An apparently straightforward answer to questions regarding the scope and limits of the psychology of language in language learning and teaching studies is that its parameters are already indicated, negatively, by the fact that it is not the only discipline that informs applied linguistics. As we have noted before, up to the mid- 1980s there had traditionally also been acknowledgement of potential inputs here from linguistics and, at times, educational theory. Positively, however, one may say that there are also inherently set limits and bounds that determine what a psychology of language can do. As with the work surveyed above, the debate within applied linguistics regarding its scientific status is particularly significant between the mid- 1970s and the mid-1980s, and embodies a discussion whose relevance has endured. So, after looking at one other early attempt at giving some kind of order to the multitude of studies, that of McDonough (1977), we will consider whether this is an adequate rendering of the scope of the psychology of language in applied linguistics, in terms of foundational criteria that aim to identify such limits and bounds.
McDonough (1977) is a good example of how early in the development of the investigation into how language is learned researchers attempted to survey and synthesize theoretical and experimental insight. Taking a cue from suggestions offered by Glaser (1976), McDonough (1977) starts by defining the role of psychology in applied linguistics in terms of several parameters. First, there is the analysis of “competent performance”. This consists of the processes and strategies of language perception and production, which categorized studies such as those by Tarone (1981) and Faerch and Kasper (1980; cf. too Cook 1977) and the work on concepts describing the initiation and production of utterances as in the monitor model (Krashen 1978: 2, 1980). Since these processes and strategies are often assumed to depend on the cognitive capacities of the learner, they must be explained by a description also of the development of competent performance (cf. McDonough 1981: 6; Glaser 1976: 18). There is no doubt that the early study of these mental capacities, mechanisms and states in learners received a new stimulus from the work of Chomsky (cf. McDonough 1981: 5, 98 f.). Chomsky’s innateness hypothesis inspired work on the acquisition of language leading, for example, to generative grammars being devised for various stages in the child’s acquisition of negation (cf. Clark and Clark 1977: 348 ff. and the discussion and references there) and other structures. Whether or not research attempted to give such an explicit characterization to the ‘innate’ language ability of learners, Chomsky’s work has continued to exert influence even in approaches that are not classifiable strictly within the men- talist/anti-mentalist dichotomy (cf., e.g., the terms in which the interlanguage hypothesis is phrased - Selinker 1972: esp. 228 ff.; Skehan 2008).
A second category of psychological research that is not strictly psycholinguistic, as the above, focuses on, amongst other things, the ‘external’ factors (in terms of conceptually restrictive linguistic criteria) that have allowed the field to be known also by the broader label of psychology of language. As part of the description of the “initial state” of the learner (cf. Cook 1985), this category (McDonough 1977: 69, 73 ff.) includes the investigation of attitude (cf. Schumann 1978; Wesche 1979; Bialystok and Frohlich s.d.), motivation (cf., e.g. Clement et al. 1977; Dornyei 2001, 2005, 2010; Dornyei and Csizer 1998), learning styles and other social psychological variables (Dornyei 2011; Dornyei and Kormos 2000). In the investigation of learning styles, one already has an indication of a disenchantment with designs for language teaching that are circumscribed by method, especially by a method of language teaching purportedly as rigorous as the audio-lingual. Instead, individual learner variables and preferences, especially learning styles, are brought into play (also see below, the first section of Chap. 7, on “A field in search of a new paradigm”).
A third component of psychological (nvestigation discussed in McDonough’s (1977) attempted synthesis is related to the pedagogical environment and the learner’s efforts in this environment, i.e. the “conditions that foster the acquisition of competence”. One may think here of the investigation of the effects of pedagogical technique, and the claims regarding group work (cf., e.g., Turner 1977; Cole 1970; Davis 1969; Rowlands 1972; Long 1975; Rogers 1978), classroom language analysis (e.g. Allwright 1980; Larsen-Freeman 1980( , the work on optimal (anguage learning environments (Burt and Dulay 1981), or the highly individualized enquiries of Hosenfeld (1974, 1976) which emphasized that ‘general’ laws of learning must be distinguished from the idiosyncratic ways in which learners actually carry out the task of language learning . A fourth category of research comprises the assessment of the process and product of language instruction: cf. here the debate on integrative versus discrete-point testing (Oller 1973; Oller and Richards 1973; Farhady 1979), and measures of success, including the description of teaching (cf. Fanselow 1977; also Voss 1984; White and Lightbown 1984).
This apparently unproblematic ordering and categorization of early and subsequent research interests in the field runs into difficulties, however, when we go beyond a consideration of what is being accounted for to ask again what might be expected of the results of such research. If the purpose is one of utility, i.e. to supply information for teaching purposes (McDonough 1981: 1), this still leaves the question of what kind of utility and what type of information we might expect. Is the utility a prescriptive one, or does it assume the form of an ex post facto justification of established practice? That it is not always easy to tell which kind it is, is clear from a number of facts. For instance, there is almost inevitably a lag between linguistic theory, upon which, amongst other things, the analysis of what McDonough (1981: 6) calls the “development of competent performance” crucially depends, and language learning research. In this way many ‘facts’ of language acquisition and perception that were explained by psychological experimentation in terms of earlier (standard or extended standard) versions of transformational grammar, may no longer need explanation, if newer versions (e.g. the revised extended standard theory, right through to minimalism) are accepted. Closer to home, perhaps, is the time difference between language learning research and the adoption of teaching styles such as the communicative approach (cf. Hatch 1978; Ritchie 1978; Cook 1981a, 1982), that we shall be returning to below. Can one possibly argue that proposals for communicative teaching should have been held back until such time as the psychology of language learning and teaching had predicted or prescribed that this was indeed the course to take, or should we rather conclude that the utility of the psychology of language is not that it is prescriptive, but rather that it allows for the subsequent justification (or rejection) of established practice?
What this brief survey of early psycholinguistic investigation makes clear is that applied linguistic reflection, of which psychology of language forms a part, does not necessarily precede, but follows professional practice, for example in giving a clarification of a problem (Crystal 1981: 7). Moreover, the assumed dependence of language teaching on psychological theory is, as we have noted before, slightly more complicated than is assumed in general statements on how, for example, the audiolingual method was based on behaviorist theory; indeed, the link between the two appears to be more tenuous and vague than is often supposed (cf. Ingram 1975: 281 f.; McDonough 1981: 9 f.). It is ironic, too, for those who would argue for a pre- scriptivist and modernist view that, where the autonomy of second language learning research is sometimes proclaimed, this appears to be necessary for defining its field of operation, even though this may have the attendant danger of isolating it from possibilities of application (Cook 1981a). It is significant, for example, that in the work of Selinker (1972: 224f.) the autonomy of the psychology of second language learning is such that no necessary connection is claimed to exist between its theoretically relevant units and the units of linguistic theory. At the same time there is the rather strict opposition of a psychology of second language learning and one of second language teaching, the latter being defined as a psychology of language learning exclusively in terms of success. But it is this last kind of theory that would have prescriptive utility for language teaching, and so obtaining autonomy for the field has its price.
There can be no doubt that Glaser’s expectations are of a prescriptive kind (cf. Glaser 1976: 7 for quite explicit claims in this regard). The problem that there are competing theories in each of the components that are described is taken care of by experimentation and re-experimentation (Glaser 1976: 8, 18), even though the prescriptive result might obscure crucial differences: in the modernist frame of mind, this is the price that has to be paid for progress.
In the work of practicing (applied) psychologists, however, expectations have always tended to be much more sober. Early claims make the point that psychology can offer no recipe for, and should exert no undue influence on language teaching (McDonough 1981: 2; cf. also the conclusion of Mackey 1973: 255 in this regard), but may provide pointers and suggestions as to more reasonable and perhaps also more preferable ways (cf. the sample of such suggestions offered by Dulay et al. 1982: 261 ff.). Then, as now, this position seems to be more in accord with the factual state of the science, discussed above. It does present a problem, however, if the experimental method is upheld as the only way of getting at the correct psychological information (cf. Glaser 1976: 8), with an appeal to the supposedly accurate, explicit and controlled research methodologies of the (natural) sciences. Then we have no guarantee against ‘undue’ influence being exerted upon teaching practice, for underlying this preference there is the expectation that, given strict control, replicability and so forth, we may yet get at the truth. We cannot ignore, however, that the intensifying debate on the foundations of theoretical reflection during the two decades that preceded this research originated in physics, one of the most ‘natural’ of natural and purportedly exact sciences. It is a fallacy, as Robins (1967 : 3) has remarked, to think that the facts and the truth are laid down in advance, awaiting discovery.
This is not to say that there is no place for experimentation or controlled observation in the psychology of language and language learning - quite the contrary, for it has an important role - but simply that it is a fallacy to believe that experiments lead to discoveries. To say that the “processes that ... speakers go through ... can only be discovered through experimental research” (McDonough 1981: 4; emphasis added) ignores the possibility that discoveries may come as a result of new hypotheses or shifts in psychological paradigms, which are only subsequently tested by experiment. Experiments are therefore set up to attempt to verify theoretical and psychological conceptualization, but one should not expect them to do more than this. A good example of this is the presentation of the hypothesis and the suggestions and guidelines for the collection of data, as well as the eventual experimental validation of the interlanguage hypothesis; cf. the statement (Selinker 1972: 222) that
the major justification .. .for writing about the construct ‘fossilization’ at this stage is that
the knowledge about ILs which turns out to suggest predictions verifiable in meaningful
performance situations, leads the way to a systematic collection of the relevant data
There is always the problem, moreover, that, however cautious the researcher may be, the experiment will prove what it set out to prove; and where experiments go wrong or do not work, explanations under a competing hypothesis may be offered (Economist 2013). This, too, may in turn prompt us to be cautious about entertaining inflated expectations about what the experimental method or theoretical knowledge can yield. Even though it may be harder to develop experimental techniques for more complex ‘alternative’ or synthetic approaches, such as interac- tionist psychology (cf. McLaughlin 1980 : 346 f.), that allowed us to re-interpret older theories from a new perspective (cf. Cook 1981b), these might make a significant contribution to our theoretical understanding by virtue of this fact. Interactionism and constructivism, for example, both constituted better theoretical justifications for the communicative designs for language teaching that were at that time beginning to emerge than any of the then current psycholinguistic investigations.
It is clear that what is needed is a way of characterizing the kind of reflection that accompanies and directs applied linguistic and psychological concept- formation, on the one hand, and that which, on the other, underlies the practical ideas that direct professional activities, such as teaching (cf. Hammerness et al. 2005). There is a hint as to the difference in McDonough (1981: 4f.): experimentation in the psychology of language is “deliberately uncomplicated” and abstract, whereas the teacher’s job is done not in terms of isolated factors but in the midst of the concrete and complicated reality of the language classroom, with all its conflicting influences (Foster 1998; cf. also Rosen 1982: 49). Supposing, in other words, that our experience is structured in such a way that when we practise theory, analytical abstraction leads the way, but when we are engage in an activity such as teaching, an other than theoretical mode of experience, which we may presume to be the pedagogical or formative mode of exercising control, guides our endeavors, then we have a set of preliminary criteria for distinguishing the two. The structure of our experience itself blocks the possibility of a prescriptivist use of science in a-scientific activity. This difference between theoretical psychological reflection and practical pedagogical tasks is evident in the sensitivity that psychologists display regarding the implementation of their research results: a psychological analysis may reveal, for example, that certain strategies and processes exist in l anguage learning, but pedagogical considerations may dictate that these not be taught (McDonough 1977: 71; Glaser 1976: 12) or that, even in principle, they might not be amenable to control in the teaching process. A possible example of the former eventuality may be simple perceptual strategies like “pay attention to the ends of words” (cf. Tarone 1981: 291), that might perhaps more profitably be inculcated by ‘guided’ instruc?tion (if at all) than by explicit teaching. Similarly, early psycholinguistic work already acknowledged that “optimal language learning strategies will not always be some simple breakdown into teachable subcomponents and the principle of grading” (McDonough 1977: 73).
Psychology of language appears most useful, then, if the kind of information it yields makes the teacher and designer of language teaching less prone to become a victim of theory. This may be true, for example, in the case of minority groups, such as the Dutch immigrant community in Canada with their Christian school movement, who sought to place their own characteristic stamp on teaching (cf. De Graaff and Olthuis 1973; Olthuis 1979) in the face of a monolithically organized national education administration. It applies equally to the other source disciplines in applied linguistics, and will be returned to again below. The psychology of language in applied linguistics will be most informative, in other words, when it probes the theoretical roots that underlie, say, competing empiricist, rationalist, interactionist, constructivist, or emergentist theories in the field. It will be least informative when, after a series of experiments, it simply prescribes the path of educational reform, as Glaser (1976: 7f.) envisaged. It is one thing, therefore, to say that psychological reflection can make a contribution to an understanding of teaching, albeit a theoretically limited one, but quite another to claim that it should effect changes in teaching: it is the teacher who is, ultimately, responsible for the decisions that are taken. Reciprocity of theoretical reflection and deliberate action (Hult 2010a, b: 22 f.) should not be mistaken for a license for prescription by the former (cf. Rosen 1982: 57).