Applied Linguistics and Language Teacher Education: What We Know
A COGNITIVE VIEW OF KNOWLEDGE TRANSFER AND USE
But why do local constraints pose such a huge problem? Why do we humans have such a problem with knowledge transfer in cases where the connection between knowledge and the new activity is not painfully clear? If we know understand what a pragmatic mistake is, why is it such a problem to notice learners’ pragmatic mistakes? Why can’t we just figure it out? The main problem seems to be what Bruer (1993) calls the “cognitive bottleneck”. Explicitly working out the grammatical, lexical and pragmatic problems individual utterances might contain, especially in real time activities with learners, necessitates a considerable amount of cognitive computing resources in short term memory (Salthouse, 1991). However, our short-term memory is very limited in the amount it can compute at once, so this kind of explicit knowledge use can lead to “cognitive overload” where the demand for cognitive capacity far outstrips the supply (Sweller, 1988). Thus, folk conceptions about knowledge use as step-by-step explicit application of propositional knowledge refer to a very inefficient use of our cognitive resources (Strauss, 2001).
Studies of expertise have shown that humans deal with this “cognitive bottleneck” by acquiring large amounts of domain specific information, especially schemata, about the activities they engage in. This allows people to simply recognize important information, which does not require much working memory capacity, rather than calculating what information is important and what it means. Chess players recognize weaknesses in opponents’ strategies, doctors notice what clinical information is important for a diagnosis; all without the need for explicitly working through the situation (de Groot, 1965; Chase & Simon, 1973; Boshuizen & Schmidt, 1992; Patel & Groen, 1991). However, it is becoming increasingly apparent that the field of applied linguistics does not represent the kind of domain specific knowledge that language teachers in order to avoid the “cognitive bottleneck” when teaching languages. (See Andrews, 2001, for discussion.)
Schön refers to this ability to understand situations and come up with possible courses of action without explicit reasoning as “knowing-in-action”. One of Schön’s major contributions to the field of education was the recognition that when professionals react in their practice without explicit, conscious reasoning, such as a teacher deciding how to respond to a student utterance, this is not some kind of automatic, Pavlovian response to stimulus. Instead these kinds of reactions represent complex, dynamic cognition and it is this kind of cognition, not explicit reasoning using declarative knowledge, which fuels practitioners’ practice (Schön, 1983).
Furthermore, Schön claims that it is principally through “reflection-in-action”, or figuring out what one is doing while engaged in an activity, that one acquires this type of “knowledge-in-action”. According to Schön, “reflection-on-action”, or thinking about what one did during an activity, has a secondary role in helping prepare people for “reflection-in-action”, but does not play a direct role in the acquisition of professional knowledge (Schön, 1987). Despite criticism of the data Schön uses to make for his claims (e.g. Eraut, 1995), others have similar explanations of how expertise is acquired. For example, after their exhaustive summary of research on expert knowledge, Bereiter and Scardamalia (1993) conclude that expertise is gained by solving both problems of procedure and problems of understanding within that particular practice. “There is no magic to how expert knowledge is acquired, but it is not enough to say that it comes about through study, experience and practice. Those terms explain mediocrity as well as expertise...problem solving provides the dynamic element in the growth of all kinds of expert knowledge.” (74)
The work of Boshuizen and Schmidt, however, show that explicit, declarative knowledge can be useful in the development of expertise. They were interested why, in studies of the process of making a diagnosis, experienced doctors did not seem to be using their biomedical knowledge (knowledge of how bodily processes work and the effect of disease and injury on these processes). Instead, the doctors referred almost exclusively to the clinical information (information which can be obtained by examining and interviewing the patient) (Patel, Evans & Groen, 1989). Boshuizen and Schmidt showed that as doctors gained robust schemata for medical diagnosis, their biomedical knowledge became associated and subsumed under their schemata for clinical knowledge, a process they call “knowledge encapsulation” (Boshuizen & Schmidt, 1992). Thus, when making a routine diagnosis, doctors only process clinical information directly, which results in less information to process and is thus cognitively efficient. However, in non-routine cases their biomedical knowledge is available to help with the diagnosis in conjunction with the clinical knowledge. While this may work well in a domain such as medicine where cause and effect are relatively clear and easy to assess, it remains to be seen whether knowledge encapsulation can work in the field of teaching where cause and effect are not as clear and are more difficult to assess.
Another strategy that people use in complex situations is to use rough “rules of thumb” which, although perhaps not always perfectly accurate, are highly cognitively efficient (Gigerenzer & Todd, 1999). Thus, a teacher who treat all mistakes as showing lack of linguistic knowledge by the language learner in situations where it is not instantly clear if the mistake was due to a momentary lapse or a deeper lack of knowledge about the L2 could be seen as efficiently using her cognitive resources.
An important part of this kind of expertise is schemata of what information is task-relevant and what information is task-redundant for a particular task or practice. Only attending to relevant information reduces the amount of information that needs to be processed, making it more cognitively efficient (Haider & Frensch, 1996; 1999). When people use these schemata to filter out unnecessary information, it greatly reduces the number of details they pay attention to and is therefore cognitively efficient (Patel & Groen, 1991). This may be one reason why, as mentioned earlier, similarity between practice tasks and target practices is so important for transfer; the more similar the knowledge is, the easier it is to recognize and use relevant information. Unfotunately, studies of teachers’ knowledge have shown that teachers’ domain specific knowledge, although similar, is significantly distinct from academics’ knowledge in the same area (Bromme, 1992; Sato & Kleinsasser, 1999; Tamir, 1992; Wilson, Shulman, & Richert, 1987). In addition, other studies have found that teachers often do not get knowledge in teacher education programs that are specific to their sphere of professional practice (Ball, 1990; NCRTL, 1991; Simon, 1993), although it must be noted that none of these studies looked at language teacher education.
A key aspect of any kind of knowledge teachers might have is awareness of constraints and affordances of a particular concept or situation (Greeno, et al, 1993; Duffy & Roehler, 1986). Constraints refer to factors inhibiting certain possible actions or states. While it may be helpful for teachers to know the difference between using open-ended vs. closed-ended questions in the classroom, they also need to be aware of the constraints of using either one. For example, in some classrooms using open-ended questions may result in problems in classroom management, especially for novice teachers, since individual answers to open-ended questions may occupy the teacher for longer periods of time, making it easier for the other students in the class to engage in disruptive behavior. The term “affordances” may seem a bit odd, but the idea behind it is fairly simple. Affordances refer to the opportunities that a certain idea or situation opens up for a teacher or what is possible to do in a particular situation or when using a specific idea. For example, many teachers do not think that grammar lessons provide opportunities to work on communicative skills. However, there are some types of grammar activities which also “affords” work on communicative competence. Inductive grammar activities where students analyze samples of language in groups, discuss patterns of form, meaning and use, and try to reach consensus on that aspect of grammar can provide plenty of communicative practice, negotiation of meaning, etc. as long as they are done in the target language. (For more in-depth discussion of affordances and constraints see Greeno et al, 1993).
Furthermore, the cognitive tasks used in different practices such as analyzing information and determining its validity are not universal, but rather different tasks are used for different practices (Amsel, Langer & Loutzenhiser, 1991; Bartels, 2003; Donald, 1995; Hativa, 1995; Brandes & Seixas, 1998; Patel, Groen & Arocha, 1990). Thus, it cannot be assumed that those cognitive tasks underlie applied linguists’ practices such as the methodological analysis SLA studies (Gass, 1995) or testing “the validity of [teachers’] principles against the observed actualities of classroom practice_ (Widdowson, 1990: 25) of will be useful or cognitively efficient for L2 teachers. For example, there is evidence that language teachers and language researchers had starkly divergent ways of validating information and have different ways of using and incorporating information into their professional knowledge (Bartels, 2003).
Knowledge organization is also important for learning a particular practice such as teaching. Expert knowledge is highly organized around the tasks people engage in, meaning experts not only recognize important information in their environment, but also that this recognition triggers possible explanations, actions and options for that specific situation (Chi, Feltovich & Glaser, 1981; Lesgold, 1984; Lesgold, Rubinson, Feltovich, Glaser, Klopfer, & Wang, 1988; Leinhardt & Smith, 1985; Schempp, Tan, Manross & Fincher, 1998; Carter, Sabers, Cushing, Pinnegar & Berliner, 1987). Because explanations, actions and options do not have to be explicitly worked out in working memory, this allows for very efficient use of cognitive resources. Studies have shown that experienced teachers’ knowledge is also organized around practice of teaching (Anders, 1995, Leinhardt & Greeno, 1986; Leinhardt & Smith, 1985; Tamir, 1992), with the linking of teachers’ mental models of the curriculum and the lesson to classroom rountines and possible explanations and representations of the subject matter (Leinhardt, Putnam, Stein, & Baxter, 1991). Studies of teacher education indicate that teacher education programs do not provide teachers with the kinds of educational experiences which would help them organize their knowledge in ways similar to experienced teachers (Carpenter, Fennema, Peterson & Carey, 1988; Calderhead & Shorrock, 1997; Gess-Newsome & Lederman, 1993; Grossman & Richert, 1988; Morine-Dershimer, 1989).
1) The provision of propositional knowledge about language can be successful in changing conceptions of and intentions for language teaching. One reason for this may be that the types of activities used in applied linguistics classes are very similar to the activity of talking about conceptions of language and language learning, thus making transfer cognitively simple.
2) The acquisition of KAL and the changing of conceptions of teaching alone does not appear to allow full and consistent transfer of KAL to L2 teaching. . The teachers in these studies did not engage in deliberate practice involving using their KAL to solve common problems of teaching practice. Perhaps while learning to teach they focused on problems of procedure (how to do things), but not on problems of understanding.
3) Well-formed KAL does not seem to be necessary to be a superior language teacher. It could be that the type of KAL investigated is not key to engaging in the practices of language teaching. It is also possible that the teachers have situation specific “rules of thumb” which are as effective as complex KAL.
4) Situational constraints pose significant problem with transferring KAL. The teachers in these studies may have lacked the classroom-based schemata and extensive mental models of using the KAL concepts in the classroom. This would then require them to do a lot of explicit processing which would result in cognitive overload.
5) Knowledge of context based instruction (CBI) was easier to transfer to planning grammar lessons than it was to use knowledge of grammar in planning CBI lessons: It would seem to be more cognitively simple to find one context for a particular grammar point and more cognitively complex to notice the great number of linguistic needs to complete a particular content task.
6) The factors that help knowledge transfer are: (a) concrete information (vs. abstract), (b) a focus on using the KAL on specific teaching activities, (c) time spent on such practice tasks, (c) deliberate practice, (d) well developed mental models, and (c) the cohesion of the teacher education program. All of these factors help teachers develop the kind of practice-specific, well-organized knowledge which reduces the cognitive complexity of using that knowledge for solving problems of practice in specific professional situations.
