Briefly, I argue that studies claiming to show a positive effect for the study of grammar typically show a small effect for grammar teaching on tests in which grammar use is possible (there is plenty of time) and encouraged (students are focused on form) and that are given a short time after the instruction. In all studies, students practice the rules extensively (Krashen, 1995, 2003). In my view, these students show only that grammar study makes a very peripheral contribution to second language competence. In addition, Haver does not cite the many research studies showing a superiority of comprehensible-input based classes over traditional classes, at both the beginning and intermediate levels (Krashen, 2003).

Haver's Approach

Haver's approach to teaching ESL is a mixture of nearly every teaching technique that has ever been used. Despite her apparent rejection of the principle of comprehensible input, she recommends the use of Total Physical Response, Natural Approach and some aspects of sheltered subject matter teaching methodology, all based on or at least highly compatible with the Input Hypothesis (Krashen, 1991). She specifically allows for a silent period and periods of reduced output. In fact, her suggested activities in chapter two of Structured English Immersion are nearly identical to those suggested in Krashen and Terrell (1983), chapter four.

After this, however, comes a deluge of techniques: Error correction of speech (p. 37), error correction of writing on the board, expansions, students' reading aloud after the teacher, students reading aloud to each other, sounding out words, sounding out words backwards (p. 46), identifying parts of speech (p. 46; see especially pp. 61-62 in which students are required to draw lines under the prepositional phrase, put a box around the verb and circle the subject), dictation (p. 46), constructing or retelling stories based on a word list (p. 47), journal writing (p. 48), sentence-combining (p. 48), transposing stories from present to past tense and vice-versa (p. 64), etc..etc. We are provided with little evidence that these activities are of value, primarily the author's word that she has used them and thinks they are worthwhile. There is great value in experienced teachers informing the field about what has appeared to be effective for them in their classes; it is important, however, to point out that research has cast serious doubt on the efficacy of many of the acttivites Haver recommends: For research on the limits of correction, for example, see Krashen, 1994; Truscott, 1996. For studies showing the lack of effect of output activities and the limits of grammar teaching, see Krashen (1994, 2003).¹

The one exception to her failure to provide empirical evidence for the recommended techniques is dictation: Haver cites Morris and Tremblay (2002) as evidence supporting the efficacy of using dictation (p. 46). Morris and Tremblay's students were students of English as a second language in Quebec, and took dictations and corrected them in groups one hour a week for 15 weeks. The dictations were actually cloze tests with unstressed function words deleted. Morris and Tremblay claim "stunning" gains on a writing test given at the end of the semester. A close look shows that the gains were not stunning at all: On the third person singular, students improved from 43% correct to 64% correct, a modest improvement (from about a D to a C). No teacher would be pleased to see these kinds of results after 15 hours of intensive practice. (Morris and Tremblay found a high level of statistical significance because their comparison group made no gains at all, which they attempt to convince the reader is normal for ESL students.) In addition, the test used was written, and students had been focused on these forms all semester, factors that maximize the amount of conscious grammatical knowledge used (Krashen, 2003).

Haver's Personal Theory of Language Acquisition

Haver does not specifically state how she thinks language is acquired or how the activities she recommends build competence, except to say that she thinks that children learn language best "when they are integrated with other children who speak that language with native-speaker ability" (p. 1). This explanation does not tell what children need to do when they are together in insure language acquisition occurs. It does not tell us how language is acquired.

2