of the learner may seem almost banal. However, this now means that the definition of the learner's needs must be expanded. First, it means that while a target language and target situation analysis must be carried out, these analyses are not enough. The EST syllabus and materials must also be based on the strategies that the learners have and will need to have (p. 230).
Kroll (1991) likewise notes that "the goal of the teacher should be to expose students to a variety of strategies... and to encourage each student to try to discover which strategies (in which circumstances) work best for him or her" (p. 252).
Larson-Freeman (1991), suggesting that grammar should be taught going from simple structures to complex ones, remarks that "what constitutes simplicity and complexity has not ever been operationally defined, although some relevant research is currently being conducted which suggests that learnability, and thus teachability, is determined by the complexity of speech-processing strategies required for particular structures" (p. 191). One of the problems in the selection of a learning strategy by either learner or teacher, is that developmental sequences of grammatical structure have not been definitely established, nor has it been established that they are identical for all learners. As a consequence, any cognitive approach is at best a trial-and-error process, at worst an educated guessing-game.
Another handicap may lie in the definition of "strategy". "A planned means to achieve change", suggest Shafritz, Koeppe, and Soper (1988, p. 417)--a definition as good as any, though rather terse. To what extent is "planning" a conscious process, even a cognitive process? And how can we classify strategies?
Wenden (1983, 1987), Rubin (1981), Oxford (1985), and others have attempted to build classification schemes. Oxford, for example, has suggested two broad categories: primary strategies (inferencing, mnemonics, summarizing, an...