So far we have pursued, albeit rather sketchily, a macro (i.e., cross-learner) and a micro (i.e., within-learner) perspective on fossilization. We have thereby established the argument that maturational constraints and native language influence are the major determinants of the general lack of success in adult L2 learning, but that the degree of such lack of success may vary from individual to individual due to the functioning of other variables. In this chapter, we will delve into one such variable, second language instruction, and explore its relationship to fossilization.

The reader may, understandably, query whether or not such an attempt makes any sense, since, in the conviction of many, one capability of instruction is precisely that it can prevent fossilization. R. Ellis (1988), for example, has advanced two claims in favor of formal instruction:

Both claims are intuitively appealing. Nevertheless, it is important to point out that while R. Ellis's second claim has by far been quite well supported by empirical evidence, his first assertion regarding instruction and fossilization has as yet received little direct investigation, theoretical or empirical. As such, it essentially remains a speculation rather than a statement of an observed fact. Lack of validation notwithstanding, the influence of the assumption has in fact been so strong, not only among some second language researchers but among second language teachers in general, that it created an if-then type of conception; that is, if there is no grammar instruction (including error correction), fossilization will result (see, e.g., Higgs & Clifford, 1982).