Why is 'research' in ELT/TEFL/TESOL/AL/SLA so irrelevant?
by Charles Jannuzi, University of Fukui, Japan
While most of the research in support of and about ELT is produced in academia, most actual language teaching and language learning are done outside academia. Even when FLs get taught at universities, the people who often end up with the teaching duties are not in the sort of university posts that are meant for research.
However, I believe the single biggest issue is the institutional approach to science and knowledge which falls under the overwhelming intellectual influences of the past half century. In short, research that is supposed to be in support of ELT is largely irrelevant, invalid and not applicable to teaching and learning because of two academic traditions: structuralism and behaviourism.
Structuralism comprises the linguistic framework for much of ELT, and yet structuralism is an historic relic of linguistics. The linguistic 'units' and 'models' and 'key concepts' of ELT are largely based on structuralist ones (and rather simplistic notions of the structuralist concepts at that). This tends to have a stultifying effect on research because it is considered bad form to question or otherwise problematize the inadequacies of ELT's simplistic versions of structuralist concepts.
The other stultifying inheritance is behaviourism (and little surprise then that most structuralists operated under behaviourist assumptions). Most research in academic SLA (an emergent field from AL that has mostly nothing to do with linguistics now, ironically enough), for example, is based on basically behaviourist preconcpetions about how to elicit 'language learning behaviour' (e.g., mastery of a 'form') from a study's subjects. The interpretation of the results (such as they are) become even more muddled because the researchers typically are not clear about whether they are looking for a psychological/psycho-linguistic OR a sociological insight about the particular population they are using as subjects (often without specifying just what population it was that they were supposed to have sampled in doing their research).
Other deficiencies abound and glare out at the teacher attempting to use SLA research to inform teaching: Very small subject groups, lack of clarity over linguistic concepts that are supposed to underly the research, inappropriate use of statistics, un-normed populations, etc. And yet an often-read, automatic defense of such research is that it, unlike qualitative research, is GENERALIZABLE. Which is simply to beg the question, because such apologists never specify to which population they want to generalize the findings. I would add, however, that I know of absolutely no finding from SLA research that is generalizable to my students, let alone one that generalizes to all of the human race now learning a FL. Moreover, it all gets even more muddled if you go back to the objection over psychological/pscyho-linguistic/cognitive goals vs. sociological ones (in which case, for example, complex differences across cultures, age groups, gender, social class and economic background, etc. become very important).
Most 'research' is done by individuals, groups and networks of people in academic posts. About the only time classroom teachers engage in such activity is when they go back to do a master's or doctoral degree under such academics.
Finally, perhaps the issue can be expanded to the following concerns:
(1) Within a socially delimited field such as 'ELT', what constitutes knowledge and who has the right to claim it as such?
(2) Are experimental and statistical procedures (mostly derived from the field of education's understanding of positivism, empiricism and probability) written up in academic journal studies the most appropriate for developing ELT/FLT and LL in most institutional settings worldwide?
(3) Much of SLA research (like the larger field of 'educational research' from which its methods came) seems to be trying to deal scientifically and experimentally with what are pscyhological, psycholinguistic and phenomenological data, and yet this line of academic research has latched onto statistical analysis of group behaviour. In short, what they are doing is trying to sociologize the psychology and phenomenology of second/foreign language acqusition, but their assumptions and conclusions are for the most part behaviouristic. Moreover, they are quite likely attempting to objectify and quantify phenomena for which parametric and non-parametric statistics were not designed.
However, the ideological apparatus/profession that is 'global ELT' is not really set up to address more than a mere handful of actual linguistic or psycholinguistic issues, and even those are treated quite arbitrarily and superficially. Add in the disconnect between the academic discourse, graduate training, materials publishing, language policy in programs and institutions (where decisions are usually made by people who know next to nothing about LT or LL), and actual classroom teaching and you get something like the world as we know it.