The Language Police: How Pressure Groups Restrict What Students Learn

Book by Ravitch Diane

One: Forbidden Topics, Forbidden Words

The greatest dangers to liberty lurk in insidious en- croachment by men of zeal, well-meaning but without understanding.
—Justice Louis D. Brandeis

I decided to write this book as a way of solving a mystery. After many years of studying the history of education and writing about the politics of education, I discovered some things that shocked me. Almost by accident, I stumbled upon an elaborate, well-established protocol of beneficent censorship, quietly endorsed and broadly implemented by textbook publishers, testing agencies, states, and the federal government. I did not learn about this state of affairs in one fell swoop, but one step at a time. Like others who are involved in education, be they parents or teachers or administrators or journalists or scholars, I had always assumed that textbooks were based on careful research and designed to help children learn something valuable. I thought that tests were designed to assess whether they had learned it. What I did not realize was that educational materials are now governed by an intricate set of rules to screen out language and topics that might be considered controversial or offensive. Some of this censorship is trivial, some is ludicrous, and some is breathtaking in its power to dumb down what children learn in school.

Initially these practices began with the intention of identifying and excluding any conscious or implicit statements of bias against African Americans, other racial or ethnic minorities, and females, whether in tests or textbooks, especially any statements that demeaned members of these groups. These efforts were entirely reasonable and justified. However, what began with admirable intentions has evolved into a surprisingly broad and increasingly bizarre policy of censorship that has gone far beyond its original scope and now excises from tests and textbooks words, images, passages, and ideas that no reasonable person would consider biased in the usual meaning of that term.

The story that I now tell began in 1997, when Bill Clinton delivered his State of the Union address. On that occasion, Clinton declared his support for national tests, and said that the states should test fourth-grade children in reading and eighth-grade children in mathematics, to make sure that they could meet national standards of proficiency. Soon after the president gave that speech, the U.S. Department of Edu- cation contracted with test publishers to develop voluntary national tests of reading and mathematics for those grades. The goal was to provide individual test scores to parents of specific children, to their teachers, and to their schools.

As someone who had been active in supporting the movement for academic standards during the 1980s and 1990s, both as a private citizen and as an assistant secretary in the U.S. Department of Education during the administration of President George H. W. Bush, I applauded Clinton’s proposal. When Bush launched his education reform initiative in 1991, he too called for national achievement testing for individual students. His plan never got off the ground, however, due to the inherently controversial nature of involving the federal government in decisions that usually belong to state and local governments; his fellow Republicans opposed it, as did the Democrats in Congress.

I supported Clinton’s program for national testing, but feared that it would falter unless it was strictly nonpartisan. If it remained under the control of political appointees in the Department of Education, it would lose credibility; whatever they did, their decisions would be criticized by members of the other party in Congress, and the testing program would come under a cloud. I made that argument in an op-ed article in the Washington Post, urging the administration to transfer responsibility for the new tests from the Department of Education to the National Assessment Governing Board (NAGB), a nonpartisan federal agency that had been supervising national testing since 1990. Why, you might wonder, was there a controversy over national test- ing if there was already a federal agency giving national tests? Let me explain.

Since 1969, the federal government has administered a test called the National Assessment of Educational Progress (known as NAEP, or “the nation’s report card”). NAEP tests are given to national and state samples of students in reading, mathematics, writing, science, history, and other academic subjects. NAEP periodically reports on the aggregate achievement of American students in the fourth, eighth, and twelfth grades, but by law it cannot measure the academic performance of any particular school district, school, or individual student. NAEP is the only regular, consistent national measure of achievement in the United States (the SAT and ACT test only college-bound students). In 1990, Congress created NAGB as a nonpartisan citizens’ board to supervise NAEP; NAGB is composed of a score of independent members, appointed by the secretary of education. NAGB has a reputation for integrity, and it seemed the right place to assign control of the new national tests that might eventually be given to millions of individual children, not just statistical samples of students. Putting the two testing programs into the same organization would also assure that the new tests proposed by Clinton for individual students would be as academically rigorous as the NAEP tests.

After my op-ed article appeared, advocating the transfer of control of the national tests, Clinton nominated me as a member of NAGB (and announced it on his weekly radio broadcast). He also accepted my suggestion to assign responsibility for his testing proposal to that board. When I joined the board at its first meeting in 1998, I discovered that Clinton’s proposed voluntary national tests (VNT) had become an important agenda item. The board spent many hours discussing the development of the new tests, trying to figure out for whom they would be voluntary (for states? for school districts? for schools? for students?), how they would relate to the established standards of NAEP, what time of year they would be given, how long they would last, how to accommodate students with special needs, whether to offer them in any language other than English, and a variety of other prickly issues.

Congress never approved the VNT. The tests were controversial from the start. Many Republicans feared that any national test commissioned by the government was the first step on a slippery slope toward federal control of education. Many Democrats objected to the emphasis on testing as opposed to new general-purpose funding. By the time Clinton left office in January 2001, his VNT proposal was dead, even though it consistently ranked high in public-opinion polls. For nearly three years, however, NAGB and the test publishers who won the federal contract worked faithfully to bring the idea to fruition, keeping a watchful eye on Congress to see whether it would eventually be authorized. It never was.

During the time that the VNT was a live possibility, the first priority was to create test questions. As a new member of the board, I was assigned to a committee that reviewed reading passages for the fourth-grade test. The committee included experienced teachers and a state superintendent of education. All of us read the passages submitted by the test contractor, a major publisher that had won a multimillion-dollar contract from the Department of Education. The committee approved passages that seemed appropriate for fourth-grade students and rejected passages that seemed dull, obscure, or incoherent. Our goal was to find short reading passages of about one to three pages, both fiction and nonfiction, written in language that was clear, vivid, and engaging, as well as test questions that gauged children’s comprehension of what they had read.

Our committee evaluated many passages for fourth-grade students. The passages had been previously published in children’s magazines or anthologies; before they reached us, they had been thoroughly vetted by the original publisher’s in-house experts. We too read them with care. As stewards of the VNT, we knew that we had to exercise extreme caution, since parents, teachers, and the media in every part of the United States would complain if anything inaccurate or untoward were to slip through unnoticed.

Most of the stories were unobjectionable; none was great literature, but for the most part, they were fairly engaging stories about children, animals, science, or history. Nearly two years later, I was surprised to learn that the passages approved by our committee had subsequently been evaluated yet again by the test contractor’s “bias and sensitivity review” panel. This panel, it turned out, recommended the elimination of several stories that we had approved. I learned that it was standard operating procedure in the educational testing industry to submit all passages and test questions to a bias and sensitivity review. Typically those who serve on these review panels are not drawn from academic fields such as English or history. Usually they have a professional background in bilingual education, diversity training, English as a second language, special education, guidance, or the education of Native Americans or other special populations. Such panels are hired by publishers, as well as by state education agencies, to screen every test and every textbook for potential bias. In the case of the voluntary national tests, the panel that scrutinized the items found biases that none of us—neither test experts nor members of NAGB—had perceived.

When publishers of textbooks and tests conduct bias and sensitivity reviews, these reviews are never released to the public; they are proprie- tary materials, and they belong to the company. I could not find a publisher willing to release them. State education departments guard the results of their bias reviews with equal zeal, even though t...