Not so fast: Some thoughts on theory culling, relativism, accepted findings and the heart and soul of SLA

This paper is meant to be a response to claims made by several prominent applied linguists in recent articles about second language acquisition (SLA) research. These claims are as follows: (1) The existence of multiple theories in SLA research is problematic (Beretta 1991), and the field should be united around a single theory or a few theories (Long 1993); (2) The alternative to such a concerted effort is a relativistic stance where 'anything goes' (Long 1990a, 1993; Beretta 1991); (3) There is now an ample body of 'accepted findings' which a good theory of SLA will have to account for (Long 1990a; Larsen Freeman and Long 1991); (4) The existence of 'accepted findings' means that SLA researchers should get on with the task of putting the findings to the test, attempting to falsify them through replication studies. I begin by disagreeing with each of these suggestions and then go on to elaborate my own view of SLA research. This view sees SLA as a process of exploration (Schumann 1993) and speculation (Davies 1991) rather than one of discovery and proof. In addition, I suggest that SLA is multi-dimensional in nature, including not only cognitive mechanisms (Long 1990a), but also the social psychology of the classroom (Allwright 1989). I end by considering how SLA research carried out according to the principles I outline might be evaluated.