Inductive and deductive reasoning are two polar aspects of the generalization process. In the case of inductive reasoning, one stores a number of specific instances and induces a general law or rule or conclusion that governs or subsumes the specific instances. Deductive reasoning is a movement from a generalization to specific instances: specific subsumed facts are inferred or deduced from a general principle. Second language learning in the "field" (natural, untutored language learning), as well as first language learning, involves a largely inductive process, in which learners must infer certain rules and meanings from all the data around them.
Classroom learning tends to rely more than it should on deductive reasoning. Traditional—especially Grammar Translation—methods have overemphasized the use of deductive reasoning in language teaching. While it may be appropriate at times to articulate a rule and then proceed to its instances, most of the evidence in communicative second language learning points to the superiority of an inductive approach to rules and generalizations. However, both inductively and deductively oriented teaching methods can be effective, depending on the goals and contexts of a particular language teaching situation.
An interesting extension of the inductive/deductive dichotomy was reported in Peters's (1981) case study of a child learning a first language. Peters pointed out that we are inclined, too often, to assume that a child's linguistic development proceeds from the parts to the whole, that is, children first learn sounds, then words, then sentences, and so forth. However, Peters's subject manifested a number of "Gestalt" characteristics, perceiving the whole before the parts. The subject demonstrated the perception of these wholes in the form of intonation patterns that appeared in his speech well before the particular words that would make up sentences. Peters cited other evidence of Gestalt learning in children and concluded that such "sentence learners" (versus "word learners") may be more common than researchers had previously assumed.
The implications of Peters's study for second language teaching are rather tantalizing. We should perhaps pay close attention to learners' production of overall, meaning-bearing intonation patterns. Wong (1986) capitalizes on just such a concept in a discussion of teaching communicative oral production.
