Note that if we ignore the Comprehension Hypothesis, that is, provide students with incomprehensible input, and force early speaking, we will raise students’ Affective Filters.

The Monitor Hypothesis is also related. The Monitor Hypothesis claims that there are severe limits to the application of consciously learned grammatical rules – learners need to know the rule (a formidable constraint) learners need to be focused on form or thinking about correction, and they need to have time to apply the rules. The only time all three conditions are met for most people is when they take a grammar test; even so, when we examine the impact of grammar study on grammar test performance, it is very modest (Krashen, 2002b). This confirms that our competence comes from comprehension of messages, not grammar study.

THE VALUE OF GRAMMAR INSTRUCTION

The Comprehension Hypothesis claims that language acquisition does not happen when we learn and practice grammar rules. Language acquisition only happens when we understand messages. This has, of course, been questioned in recent years, as a stream of papers have appeared in the professional journals claiming that grammar instruction is helpful. I am pleased that these studies are being done: What was once an axiom is now a testable hypothesis.

In my reviews of these studies, I have concluded that they confirm the correctness of the Comprehension and Monitor Hypotheses: they show only that even after substantial grammar study, even very motivated students show only modest gains in accuracy, and these gains occur only on measures that encourage a focus on form. Truscott (1998) has arrived at very similar conclusions.

Some have interpreted this position as a claim that all grammar teaching is forbidden. Not so. There are two good reasons for including grammar in the EFL curriculum.

The first is for "language appreciation," otherwise known as "linguistics." Linguistics includes language universals, language change, dialects, etc. The second is to fill gaps left by incomplete acquisition and places in which idiolects differ from the prestige dialect. Society’s standards for accuracy, especially in writing, are 100%: We are not allowed "mistakes" in punctuation, spelling or grammar. One public error, in fact, can result in humiliation. Even well-read native speakers have gaps, places where their grammatical competence differs from accepted use.

Consciously learned rules can fill some of these gaps, which are typically in aspects of language that do not affect communication of messages. The place to use this knowledge is in the editing stage of the composing process, when appealing to conscious rules will not interefere with communication.

I recommend delaying the teaching of these rules until more advanced levels. I would first give acquisition a chance, and then use conscious knowledge to fill in some of the gaps. There is no sense teaching rules for Monitoring that will eventually be acquired.

Grammar, thus, is not excluded. It is, however, no longer the star player but has only a supporting role.

2