The dream of absolute prediction—this method, under these conditions, produces this result—has been a characteristic of utopian thinking from the 18th century on.
In his whirlwind tour that accompanied the signing of the “No Child Left Behind” Act of 2001, President Bush stressed the importance of reading instruction based on “science"—not “what sounds good.” We have, of course, had presidents who were committed to educational improvement, but never one so focused on reading, even to the point of entering the Reading Wars and aligning himself with a particular research base.
The scientific principles Mr. Bush alluded to are laid out in a report from the National Reading Panel, “Teaching Children to Read,” which somewhat redundantly claims to be “an evidence-based assessment of the scientific research on reading and its implications for reading instruction.” According to the reading panel’s project director, Duane Alexander, the goal of the study is to help “ensure that reading instructional approaches in America’s classrooms reflect scientifically based methods.”
This has long been a dream of experimental educational research, to transfer the scientific methods of the hard sciences to school learning. The key question must be: Does the report deliver on its promise to provide a solid, incontrovertible base of research conclusions that can usefully guide classroom practice? Can it help schools move from “what sounds good” to “science”?
The answer can only be no. No, despite the diligent work of the panel, performed under considerable time pressure. No, because the report looks at so few areas and even in these, the research base often yields equivocal results or conclusions that are simply too general to be useful to teachers. No, because the panel chose to look at only one kind of research.
The reading report stunningly fails to find any solid evidence in support of independent reading, largely because it dismisses all correlational studies. Correlation, of course, does not demonstrate causation, but even fields like medicine and epidemiology regularly make use of it when experimentation is difficult (the effects of cigarette smoking, for example). If proficient readers typically read extensively on their own, as the research suggests, it would seem prudent, even scientific, to develop this habit in young readers. I suspect that few of the panel members themselves would want their children in programs that did not include independent reading.
But the study concludes tepidly that extensive silent reading “may” help comprehension. It excludes independent reading from its list of endorsed practices, consigning it to a kind of limbo. Joanne Yatvin, the one dissenting member on the committee, accurately predicted that the report would create suspicion about practices that either were not studied or for which the evidence is not clear-cut. It is, then, only one small step for school districts to shift resources from nonvalidated to validated practices. We can see this happening in some California schools, where teachers have been pressured to remove classroom libraries because they interfered with instruction.
Even the most definitive conclusion of the study, supporting the value of phonemic-awareness instruction for younger children, is hedged with qualifications and areas of uncertainty—white spaces that a teacher must fill. For example, the research has not determined how many months or years a phonics program should last, or how to flexibly meet the needs of different children. The report advocates an “integrated” approach to reading, with phonemic awareness a key early component, but it does not provide the teacher with a plan for that integration. As Ms. Yatvin notes in her dissent, it does not answer the key situational questions teachers must face—when, how much, and for whom?
It may be time to ask if reading science can truly deliver on its promise.
One of the cruel paradoxes of this report, which leaves so much to teacher initiative and flexible decisionmaking, is the way that it has been distorted and used as “research support” for scripted approaches like the McGraw-Hill Open Court series, a linear descendant of the rigid DISTAR program, which allowed almost no room for teacher decisionmaking—or student choice.
But teacher control is never far from the agenda of reading science, either. The dream of absolute prediction—this method, under these conditions, produces this result—has been a characteristic of utopian thinking from the 18th century on. It is, after all, the dream (or nightmare) of the Crystal Palace, that Dostoyevsky attacked in Notes from the Underground, the promise that a certain type of knowledge will allow for constructive and predictable ordering of human relationships.
We now have available almost a century of experimental research in reading, much of it designed to help educators along the road to this predictive certainty. Yet, on the questions most critical to actual teaching, this summary report must point to the promise of future research, a hopeful horizon that keeps receding each year. How long must we wait? It may be time to ask if reading science can truly deliver on its promise. And, like Dostoyevsky, we might ask if we would want to live in this ordered house, even if it could be built.
Thomas Newkirk is the director of the New Hampshire Literacy Institutes and is an acquiring editor for Heinemann Books. He teaches in the English department at the University of New Hampshire in Durham.
A version of this article appeared in the April 24, 2002 edition of Education Week as Reading and the Limits of Science