In my reviews of the research on in-school free reading (Krashen, 1993; 2001), I have concluded that with very few exceptions, students in these programs progress in reading at least as well as those in comparison groups, and often do considerably better. The most successful studies are those that last for longer than one academic year. Short-term studies produce less than spectacular results, most likely because it usually takes readers some time to settle in and find suitable reading material.

The Comprehension Hypothesis

Not only do we see triangulation among three different sources of evidence, all supporting the effectiveness of free reading, the Reading Hypothesis is also consistent with the more general Comprehension Hypothesis, the hypothesis that we acquire language by understanding it. The Comprehension Hypothesis finds support in several areas in second language acquisition, including beginning language teaching and intermediate language teaching: Students make better progress in classes that supply more "comprehensible input." I have also argued (Krashen, 1994, 2003) that other means of acquiring language are unlikely and without empirical support. Goodman (Furkey and Xu, 2003) and Smith (1994) have presented strong evidence that comprehension is the means by which we learn to read, providing evidence from sources very different from those cited here.

Applications

Sustained silent reading

Perhaps the most common application of the Reading Hypothesis in schools is the practice of sustained silent reading, or SSR. In SSR, a few minutes each day is devoted to recreational reading, usually between five and 15. As noted above, the research on SSR and related programs is quite positive. Pilgreen (2000) presents the principles of SSR and some helpful guidelines. I focus here on just a few aspects of SSR.

Do they really read during SSR?

Some critics have claimed that SSR is merely a time for "pretend" reading, that many students simply flip pages and look at pictures. Von Sprecken and Krashen (1998) observed SSR classes in a middle school in the middle of the school year and reported that 90 percent of the students were engaged in reading. Their observations were confirmed in subsequent studies (Cohen, 1999; Herda and Ramos, 2001).

Von Sprecken and Krashen also reported that more reading tended to take place in those classes in which more books were available, in which teachers read at the same time students read (see also Wheldall and Entwhistle, 1988), in which students were not required to bring their own books, and in which teachers made deliberate attempts to promote certain books.

From observing SSR classes and from talking to numerous professionals, I have concluded that is more effective to do a little each day than to devote large segments of time once or twice a week to free reading. The goal of SSR is to develop a taste for reading, to stimulate the once-reluctant reader to read more outside of school. Rather than forcing reading, and possibly making it distasteful, small doses are much more likely to work. It is not the actual time reading during SSR that counts, it is the desire to read more than counts. Thus SSR is not for very advanced readers. It is doubtful that readers of this paper will benefit by adding five to 15 minutes of free reading to their daily schedule. We have already developed a reading habit. Studies confirm that those who participate in SSR programs read more on their own when the program is over (Pilgreen and Krashen, 1993) and one study showed that they continue to read more even years later (Greaney and Clarke, 1973).

Self-selected reading

Self-selected reading programs were popular in the United States in the 1950’s. In self-selected reading, the entire class period is devoted to recreational reading, except for a small amount of time devoted to teacher-student conferences in which teachers discuss what the child read, any problems that may have come up, and recommend additional reading.

2