One can learn a great deal by surveying the physics textbooks now in widespread use, as I recently did as part of a project directed by the Environmental Literacy Council. Begin with some superficial impressions: On nearly every page one finds boxes, insets, three-dimensional marginalia in four colors, and all manner of gratuitous graphics. It is difficult to discern any rank order to the different kinds of information presented. Since physics depends on coherent argument, this manner of presentation is clearly ill-suited to the books’ purpose.
Naïvely, I initially thought the formatting of the books might be intended to suit the cognitive peculiarities of today’s students. I recently taught a course (not physics, but Latin) in a suburban public high school. I was shocked to discover that relatively few students at this “Blue Ribbon National School of Excellence” (so says the Department of Education) seemed capable of real concentration. My impression was confirmed by veteran teachers who speak of a dramatic change in students over the last fifteen years. The culprits they name are familiar enough: the near-complete demise of reading, coincident with the rise of video games and the Web. The ability to follow a monological narrative or argument from beginning to end seems to have been diminished, along with the habit and taste for reading. So surely the textbooks are adapting to this sad fact in a principled way, out of necessity, guided by the latest findings of cognitive science? Not so, it turns out — as becomes all too apparent when one learns how textbooks get produced.
Not surprisingly, the textbooks offered by publishers are products of market demand. Like any market, the market for textbooks does not exist in a political vacuum. In the U.S. there is no omnipotent ministry of education that sets standards for curriculum; the states set their own standards. There are a number of well-meaning, semi-official organizations that try to bring good sense to bear on the chaos. For example, in 1993 the American Association for the Advancement of Science issued its “Benchmarks for Science Literacy,” and in 1996 the National Research Council issued its “National Science Education Standards.” But these efforts have had little real effect. The movement toward states setting their own standards received a sort of federal blessing with the Improving America’s Schools Act of 1994, which linked federal funding for schools to states’ efforts to set content standards and assess academic performance.
The problem is not so much federalism itself as the way most states actually operate. Each jurisdiction sets out standards in excruciating detail, including long lists of topics to be covered. These lists are given to the publishers as “bid specs.” Publishers can maximize the likelihood that a book will be widely adopted by including everything in the bid specs of the ten or fifteen biggest markets. It is no surprise, then, that textbooks often run as much as 900 pages long. The reason this simplistic business strategy is successful, and the reason publishers’ salesmen push for allocating an ever greater share of a book’s development costs to graphics, has to do with the way textbooks are adopted.
The members of a state’s textbook adoption committee are often appointed so as to obtain geographical and political representation (they come from different congressional districts), not because of any expertise in the subject matter. In most states it is a one-time appointment, and a form of political patronage. Members generally have other jobs. In a typical scenario, they come away from their first meeting with a couple of documents from the state’s Curriculum Task Force to help them: a list of several hundred “behavioral objectives” to be accomplished by the state’s schooling in general, and another list of perhaps a hundred topic items specific to the subject. The coverage of these topics is to be checked off on a form, ranked on a scale of one to five for each textbook and the mass of materials that go with the book: teachers’ editions, consumable workbooks, wall charts, ready-made transparencies and exams, demonstration materials, lab manuals, and all manner of classroom pizzazz. Committee members had better have a spare bedroom available. As one close observer put it, “Back in their homes, committee members leaf through the mass of materials aimlessly, not sure of what to look for. Some members alight on pages they don’t understand. Some of them conclude that things must have changed quite a bit since they were young, and others conclude that they are too tired to tackle the task and go to bed.”
Even for a diligent committee member, the best that can be accomplished under such a system is merely to ascertain the presence of requisite topics, not the clarity or depth with which they are presented. The reality is that nobody involved in the selection process is actually reading the books, so from a publisher’s perspective, the important thing is that every conceivable topic be mentioned and, just as important, listed in the index for quick reference. The Third International Mathematics and Science Study found that the average U.S. science middle school textbook covers 50 to 65 topics, while texts in Japan include only five to 15 topics and German textbooks cover an average of seven topics. The superficial treatment of dozens of topics comes at the expense of students’ conceptual understanding.
In the end, the question whether students can get any pleasure or meaning out of the text is never really brought to bear on this process. And the superficial nature of the selection process dictates a coffee-table approach by the publishers, leading them to produce a lavish physical product that is heavy on impressive-looking graphical clutter. “Thus the de facto national curriculum is a thin stream of staccato prose winding through an excessive number of pictures, boxes and charts,” as Harriet Tyson-Bernstein puts it in A Conspiracy of Good Intentions. Teachers, of course, needn’t follow such texts slavishly, and in fact those teachers who have real mastery of their subject typically depart from the text and conduct their own classroom investigations. But textbooks are relied upon quite heavily by less-experienced and less-knowledgeable teachers. More generally, research conducted as part of the Third International Mathematics and Science Study (TIMSS) indicates that textbooks have a major impact on teachers’ decisions about how to present their subject material. In a survey of 16,000 science teachers conducted by Education Market Research in 2001, over 80 percent of science teachers reported using a traditional science textbook.
In many middle school texts, questions at the end of each chapter require students to do no more than repeat some definition verbatim from the text. Students generally have deeply held prior beliefs about natural phenomena, often wrong, and they can easily answer such textbook questions without recognizing the inconsistency between their understanding and the text they are memorizing. The result is that students work their way through the material without being changed by it, and often without really seeing the point of the questions asked or the answers given.
