Stephen Krashen University of Southern California I have several comments on Hwei-Jiun Shen's Role of Explicit Instruction in EFL/ESL Teaching (FLA, Shen, 2003, 36[3]). 1. Shen argued that California fourth graders' low scores on a national reading test (NAEP) since 1992 are evidence for the failings of literacy approaches (p. 425). This has been a popular interpretation, but there is good evidence that it is not correct. NAEP scores were analyzed by state for the first time in 1992. It was assumed that reading proficiency in California was better before progressive literacy approaches (whole language) was introduced but there is no NAEP data to support this. McQuillan (1998), however, examined California children's performance on other tests of reading from 1984 to 1990 and found no significant change, demonstrating that California was in trouble before 1992-well before whole language was introduced in 1987. McQuillan also presented compelling evidence that California's children have the worst access to books in the United States, with the worst school libraries, poor public libraries, and high rates of poverty in families with school-age children. McQuillan and others have reported clear and substantial correlations between access to print and reading achievement (see also Krashen, 2002). In addition, there was insufficient time for the decision of a committee made in 1987 to have an impact on tests in 1992: Fourth graders taking the test would have had less than one year under the new curriculum. It should also be pointed out that California instituted a skills-based approach soon after the 1992 scores were announced, and whole language was banished. The most recent NAEP results (2002) show that California still ranks at the bottom of the country and its scores have not improved since 1992. California's fanatic rush to phonics and skill building had no effect. This is a stunning confirmation of McQuillan's arguments; there has been no improvement in California's school libraries or public libraries in the last 10 years. 2. Shen concluded that first language research shows that explicit instruction (in phonemic awareness) is necessary, based on the report on the National Reading Panel. Shen may not be aware that this has also been challenged. In Krashen (2001a), I reviewed the reading panel's research on the impact of phonemic awareness training on reading comprehension: I found only 6 studies and 13 comparisons, which included only 3 studies of English-speaking children. Overall results were not spectacular, and several studies showed no effect at all. In fact, only one study produced consistently positive results, a study done in Israel with 15 Hebrew-speaking children in the experimental group. In their response to me, members of the National Reading Panel (Ehri, Shanahan, &r Nunes, 2002) commented that more studies might produce more positive results. Of course, I agree, but until that happens, we cannot claim that phonemic awareness training has a substantial impact on reading. The research literature also shows that many children with low or even no phonemic awareness learn to read quite well, and that phonemic awareness develops without instruction (Krashen, 2001b). Most likely, phonemic awareness is the result of reading, not the cause. 3. Shen stated that researchers on implicit instruction have tried to show that natural implicit instruction works well with ESL/EFL as well as with Ll learners (p. 426). Researchers have done more than try. In published study after published study, comprehensible-input based methods, when compared to traditional methodology, have been shown to be very successlul at beginning and intermediate levels, without serious exceptions (For a review, see Krashen, 2003), and extensive reading and sustained silent reading have consistently been shown to be more successful than traditional instruction in L2, as well as in Ll development (reviewed in Krashen, 2001c, 2003). …