Publications
Publications
defining ideas
policy review
china leadership monitor
education next
hoover digest

See Dick Flunk

Decades of research shows that kids with reading problems need phonics-based instruction. Why aren’t educators listening?


The evidence is overwhelming that kids
with reading problems need phonics-based instruction.
Why aren’t educators getting the message?

Inside a National Institutes of Health (NIH) reading lab, 11-year-old Alexis stumbles to decipher a short story. Reading out loud, she inserts the word "girl" at the end of a sentence in which it does not appear. She skips the word "the" and says "grader" instead of "grade." Instead of "goes," Alexis reads "got"; instead of "her," Alexis guesses "the." Later in the sentence, she substitutes "broom" when the words read "a round iron handle." This is a sobering display: Alexis, an otherwise bright sixth-grader who scores above the 70th percentile in all other academic areas, cannot read a simple sentence without several mistakes and frequent guesswork. Unfortunately, she is not alone.

Alexis is one of more than 10,000 participants in an ongoing 30-year, $200-million study of reading disabilities by the National Institute of Child Health and Human Development (NICHD) , a division of the NIH. Acting NICHD Chief Reid Lyon sadly notes that her case is typical of children who have not received proper instruction in how the sounds heard in speech are represented by the letter symbols used in print—the relationship known as phonics. Says Lyon, "There is no way to read if you are not very facile in the use of phonics."

The problem is that few readers experiencing difficulties similar to Alexis’s are ever given the explicit phonics training they so desperately need. Instead, teaching methods variously termed "look-and-say," "sight method," "whole word," and the latest incarnation, "whole language," have dominated the education landscape for almost seven decades. As a result, millions of kids are consigned to a lifetime of unnecessary reading troubles because most policymakers and educators have either willfully ignored the NIH-funded research or are unaware of its existence.

This is clearly evident in the America Reads Challenge Act of 1997, President Clinton’s five-year, $2.75-billion proposal to place volunteer reading tutors with minimal training in low-income schools. The program would hire reading specialists to give cram courses to these volunteers, but declines to incorporate the NICHD’s findings into its recommendations. Its official literature tepidly states, "The U.S. Department of Education does not specify any particular reading instruction method." In addition, the federal government gives elementary schools $7 billion a year in aid to programs for special education, bilingual education, and low-income students without insisting that the instruction be research-based.

If Clinton’s remedy is misguided, at least his focus on reading is well placed. The 1994 National Assessment of Educational Progress (NAEP) found that more than 40 percent of fourth graders cannot read at the most basic level, indicating that they could not understand the "overall meaning of the text" or make simple inferences. The 1993 National Survey of Adult Literacy discovered that some 90 million Americans—nearly half the adult population—have severely limited literacy skills, and their ranks swell by millions each year. Bereft of the ability to use a bus schedule, write a short letter to address a credit problem, or calculate their savings on a sale price, they are much more likely to be unemployed, on welfare, or in jail than their fully literate peers. More worrisome is the fact that literacy skills among young adults and school-age children are declining.

Among minorities, the statistics are even more tragic. On average, black and Hispanic children score four grade levels below their white peers on reading tests. And this gap does not narrow over time: The average black college graduate reads at the level of the average white high-school graduate. To be fair, American schoolchildren overall ranked second only to Finland on the last international assessment of reading ability, but that provides little consolation to disadvantaged children who scored well below the average score of our major trading partners. Clearly our education system is leaving too many of its most vulnerable charges far behind in an age when literacy is the gateway to most important skills.

The Reading Wars

What these kids don’t know is that they are the casualties of what has been labeled the "reading wars." Across the country, school districts are embroiled in a bitter, decades-old dispute over how best to teach reading.

The latest uproar is swirling around the controversial whole-language theory first introduced in the early 1970s. Its supporters contend that children will learn to recognize individual words through actual reading, using context, pictures, and familiar words to understand the meaning of written passages even if they can’t read every word. They deride skills-based phonics instruction as abstract and boring, favoring techniques such as reading to children and encouraging them to read and write early and often. "It’s in the interaction with the text that children develop good solid hypotheses about the text, not through segregating sounds from the text," says Sharon Murphy, the outgoing president of the Whole Language Umbrella , an independent professional association.

This approach gained thousands of acolytes during the 1980s. The nation’s colleges of education produced a new crop of teachers weaned solely on whole-language philosophy, while influential professional associations such as the National Council of Teachers of English and the International Reading Association (http://www.ira.org/) embraced its basic premises. At the state level, California spearheaded a virtual reading revolution. The state department of education rewrote its entire curriculum in 1987, ditching phonics for a literature-based, whole-language approach. Teachers were told to throw out their old methods and embrace the cutting edge. Other states and local school districts soon followed. "All the major publishers moved to whole-language readers once California implemented it," says Bonnie Grossen of the National Center to Improve the Tools of Educators, at the University of Oregon. "They had no sequenced instruction, just pretty pictures and poetry. It has taken hold in all 50 states."

Yet while educators and textbook publishers were enthusiastically welcoming whole language, the research evidence supporting phonics-first instruction and questioning the underpinnings of whole-language theory continued to mount. In 1985, the U.S. Department of Education released "Becoming a Nation of Readers," a report which concluded that "the issue is no longer . . . whether children should be taught phonics. The issues now are specific ones of just how it should be done."

Another federally funded study led to the publication in 1990 of Beginning To Read, which most researchers consider the seminal review of the pertinent scientific literature. Its author, Marilyn Adams, now a visiting scholar at the Harvard Graduate School of Education, says, "You can teach children more efficiently and effectively if you use phonics. If you don’t know how the alphabet works, you can’t learn how to use an alphabetic language. There is no argument."

These findings are beginning to have an impact. Several states, including North Carolina, Texas, Georgia, Washington, Wisconsin, Oregon, and Ohio have recently passed legislation recommending phonics education in the early grades. "We no longer will accept that kids cannot learn to read," says Cindy Cupp, the director of reading at the Georgia Department of Education. "Now the state is in favor of explicit phonics instruction."

The International Reading Association recently reversed policy, specifically promoting early phonics instruction as a necessary component of a comprehensive reading program, and the American Federation of Teachers (AFT) has come down squarely on the side of skills-based instruction for beginners. "We created a terrible nightmare for a lot of kids who haven’t been able to learn to read using whole language by itself," says Beth Bader, the assistant director for educational issues at the AFT, the nation’s second-largest teachers union. The National Education Association, the largest teachers union, still gives lukewarm support to whole language, but spokeswoman Karen Smith grudgingly admits that "many kids cannot learn to read without phonics." Dozens of news stories from school districts nationwide catalog widespread discontent with whole language and a resurgence of support for phonics-first instruction.

Most damaging to whole language’s adherents, last year California punted its whole-language curriculum altogether, stressing the need for systematic, explicit phonics instruction in the early grades. The state reversed course in response to a wave of public criticism after California’s poor performance in the 1994 NAEP, when it tied Louisiana for last place. Janet Nicholas, a member of the California State Board of Education, recently told the U.S. House Committee on Education and the Workforce: "Unfortunately for California children, the unsubstantiated claims and enthusiastic visions of whole-language ideologues proved to be disastrous when applied to real children."

The reaction to California’s actions was predictable. "Whole language is being used as a scapegoat for dropping scores, when California has many minorities and high immigration," says University of Arizona education professor Ken Goodman, regarded by many as the godfather of whole-language theory. It is true that whites are a minority in California and a large portion of its Hispanic population are recent immigrants who speak bare-bones English. Yet apologists for whole language ignore the fact that scores dropped equally among children whose parents graduated from college.

"These data [from the NAEP] underscore the fact that reading failure is a serious national problem and cannot be attributed to poverty, immigration, or the learning of English as a second language," says Reid Lyon, who has directed the NIH reading studies for the past six years.

The 10-Year Itch

Fortunately, many educators are admitting their mistakes and switching to what works. But it is easy to be skeptical about whether these reforms will last. The American education system is notorious for swinging dramatically from one philosophy to another. Embattled educators and parents looking for a quick fix rarely give meaningful changes time to work. Historically, this latest reading shakeup fits into a pattern of reform and retrenchment dating back at least a century.

During colonial times, the formula was simple: Teach kids the relationship between letters and sounds and then let them read. This method went unchallenged until the mid-1800s, when the influential educational reformer Horace Mann excoriated the drilling methods of the past. In the stark language of his reports to the Massachusetts Board of Education, the letters of the alphabet were "skeleton-shaped, bloodless, ghostly apparitions." Instead of teaching individual sound-letter relationships, Mann thought children should focus on comprehension by learning whole words first.

Despite his suggestions, through the early part of this century most American schools continued to use the traditional method of first teaching the 44 sounds heard in speech and then relating them to the 200 letters and letter groupings that appear in English. Once they had mastered these skills, it was presumed, most children could "sound out" any word, even unknown ones. Comprehension was only limited by their speaking vocabularies.

To the layman, this makes perfect sense. As education professors Connie Juel of the University of Virginia and Isabel Beck of the University of Pittsburgh write in the AFT journal American Educator, "Given that letters and sounds have systematic relationships in an alphabetic language such as English, it stands to reason that those responsible for teaching initial reading would consider telling beginners directly what those relationships are." But progressive educators based at Columbia University Teachers College and the University of Chicago in the 1920s rejected the "code-emphasis" approach as an unnatural, undemocratic way of learning. Phonics was derided as the "drill-and-kill" method, evoking images of stern nuns leading chorus recitals of "a," "oo" and "th."

These educators reintroduced Mann’s idea that children could read by learning to recognize whole words in context. Skills-based instruction, they argued, discouraged kids from acquiring a love of reading because of its rote drilling and memorization. What influential educators such as John Dewey advocated soon became known as the look-say approach. Textbook publishers responded quickly. Whereas colonial children (at least upper-class children) learned to read using Noah Webster’s bestselling Blue-Backed Speller and the Bible, mid-20th-century youngsters were subjected to the simplistic, mind-numbing "Dick and Jane" series. Responding to children’s limited capacity for memorizing whole words, school readers became increasingly repetitive and wholly uninteresting. "We stopped teaching kids rules," says Bader of the AFT, "and expected them to learn 2 million individual words instead of teaching them 100 rules to figure them out."

Look-say reigned controversy-free until 1955, when Rudolf Flesch published Why Johnny Can’t Read. Flesch, an admirer of Dewey with a doctorate from Columbia Teachers College, criticized the look-say approach in strident language: "We have decided to forget that we write with letters and learn to read English as if it were Chinese. One word after another after another after another. If we want to read materials with a vocabulary of 10,000 words, then we have to memorize 10,000 words; if we want to go to the 20,000 word range, we have to learn, one by one, 20,000 words; and so on. We have thrown 3,500 years of civilization out the window and have gone back to the age of Hammurabi."

Flesch’s critique of the education system, in which he likened current methods of reading instruction to the training of dogs, was understandably not well received. Yet his basic claim that look-say was unsupported by research piqued the curiosity of at least one noted researcher, Jeanne Chall of the Harvard Graduate School of Education. She conducted a three-year study and a massive literature review, publishing the results in Learning to Read: The Great Debate (1967), still widely read among educational researchers. Its central conclusion was that the evidence favored the code-emphasis approach, particularly for poor children and those showing early signs of dyslexia. An updated version published in 1983 surveyed the research since 1967 and found that the case for phonics-first instruction was even stronger.

These findings led to a revival of phonics-based instruction during the 1970s, only to have these gains undermined by the increasing influence of whole-language theory. Its grounding in a distinct philosophy of language and harsh indictment of the "Dick and Jane" readers distinguished it from the look-say approach, yet in the most important respect whole language did not differ at all: It defied common sense and ignored piles of research by de-emphasizing skills and focusing almost solely on comprehension. It was merely the latest fad in a long line of meaning-first approaches.

A Theory Discredited

Much of what whole-language advocates claim to have introduced is uncontroversial. All educators, including those who support phonics instruction, believe that children should begin reading real literature and writing as early as possible, that comprehension is the ultimate goal of reading, and that education should be relevant to students’ lives.

It is the philosophy undergirding whole-language theory that troubles linguists and research psychologists who study how the brain processes language. The founders of whole language set themselves apart from the look-say crowd by advancing not only a new method of reading instruction but a new theory of how children acquire written language abilities. Isolated drilling in sound-symbol relationships was unnecessary, they argued, because learning to read would be as natural as learning to talk if meaning and purpose were emphasized. Indeed, whole-language theorist Frank Smith argued that skilled readers skip around instead of reading each word, using context to confirm hypotheses about the meaning of text. Hence education professor Ken Goodman’s description of reading as a "psycholinguistic guessing game." To a whole-language disciple, phonics instruction can only take place as a rare intervention while children are actually reading.

Whole language’s infatuation with the contextual nature of reading is moored in a 1965 study by Goodman. During the study, beginning readers were given a list of words and then a passage with the same words in context. Observing that children’s word-identification skills improved after reading the passage, Goodman concluded that context plays a central role in deciphering text.

But when researcher Tom Nicholson revisited the study in the Journal of Educational Psychology in 1991, it fell apart. By controlling for the children’s reading-skill levels and the order in which they received the two tests (to eliminate the "practice effect"), Nicholson found that context only helped poor readers and offered readers in general no significant benefit. "Goodman based his ideas on a poor study whose findings were never replicated," says Lyon of the NIH. "It never would have gotten through a National Institutes of Health review."

Eye-movement studies have further undermined whole language’s faith in context by proving that skilled readers do not use context and prediction to capture a text’s meaning; they actually process each word visually. Other studies by Keith Stanovich of the University of Toronto and Charles Perfetti of the University of Pittsburgh have shown that good readers seldom rely on context; instead, their decoding skills are so practiced and quick that they speed through text without effort. Less able readers struggle painfully to identify words, taxing their ability to understand the text. "It is only because readers (and listeners) process words so automatically and effortlessly that they have the mental time and capacity left to construct and reflect on that meaning and message," write Marilyn Adams and Maggie Bruck, of Montreal’s McGill University, in American Educator.

The belief that reading is a "natural" activity entails changing the schoolhouse dramatically. Whole-language teachers tend to regard themselves as motivators rather than instructors, instilling enthusiasm instead of basic skills. For example, they favor "child-centered" over "teacher-directed" classrooms. "Children should be fully active participants in building your classroom environment and curriculum, engaged in all the critical and creative thinking those tasks require," writes Bess Altwerger, a leading whole-language proponent. "Even first-grade students are capable of working collaboratively in this regard, as long as you can accept an environment reflecting the development of children rather than adult proficiency."

These changes are troubling when one considers that Smith and Goodman’s belief that learning to read is as natural as learning to speak is "accepted by no responsible linguist, psychologist, or cognitive scientist in the research community," writes Keith Stanovich, one of the foremost reading researchers in the world. Barbara Foorman, an educational psychologist at the University of Houston and an NIH researcher, points out that if reading were as natural as speaking, there would be no illiteracy in literate societies.

Although the basic principles of whole language have been discredited, its proponents are not bending. "There are different kinds of research, qualitative and quantitative," says Murphy, "Whole-language researchers tend to fall on the qualitative side." In short, they question the research method instead of answering the research. Indeed, many whole-language proponents have resisted evaluating their approach using traditional measures of student performance, preferring such techniques as "kidwatching" and long-term evaluations of students’ "real" written work. They generally question the reliability of standardized tests and controlled studies as artificial methods that fail to take account of cultural and environmental differences. Phonics advocates regard this as subterfuge. "Some people just don’t want to rely on research, which means that we repeat the same errors over and over again." says Harvard’s Jeanne Chall with obvious frustration. "It’s very sad."

Phonics Ascendant

In a training tape developed for teachers in California, an expert teacher-trainer demonstrates phonics in action. She holds up a "very hungry" stuffed bear named "Chuck," who is "choosing lunch." His diet, however, is limited: Chuck only wants foods that begin with the same sound as his name, such as "cheese, chips, and chopped-up chunks of peach." The teacher asks her class what else Chuck might like. One first-grader correctly ventures "cherry pie." Another child, as confused about edibility as the sound "ch," mistakenly offers "jacket." After briefly explaining his error, the teacher shows her students the word "Chuck," and points to the first two letters. In unison, her engaged children practice making the sound.

This exercise helps the children realize that the letters "c" and "h" together make a familiar sound heard in many of the words they speak. Afterwards, the teacher gives her children a small book that lets them practice this new skill by including many words spelled with a "ch"—what educational researchers term "decodable text." Sure, it’s not Treasure Island, but the kids are learning to associate the sound with the symbol, enabling them to read "real" literature in the future. This is anathema to followers of whole language. "People from literature-based philosophies would freak out if they saw this. They don’t want to work with kids on these subskills," says Lyon.

It is these subskills, however, that impoverished children and those suffering from reading disabilities such as dyslexia need the most. Research suggests that direct instruction in phonics is innocuous but unnecessary for the most able 50 percent of children. Neurologists speculate that their brains may be "hard-wired" at birth to dissect speech into individual sounds and, with a little formal instruction, easily match those sounds to individual letters and syllables. Once exposed to generous helpings of language, these kids quickly move from "see Spot run" to richer literature.

The next quartile of children will learn to read, but they may fall behind without strong early phonics instruction. For the remaining 25 percent, though, reading will be one of the greatest challenges they will face in life. To enable them to meet that challenge, Lyon says, "phonics is nonnegotiable." Without systematic, explicit instruction in the sound-symbol relationships that comprise the English language, they will not read with the facility required to glean meaning from text.

The NIH studies have demonstrated this over and over, at 12 sites including Yale, Johns Hopkins, Harvard, Florida State, and the University of Houston. These studies have shown that the best predictor of the ability to comprehend text is the speed and accuracy with which a child reads individual words. In essence, good readers use phonics constantly, only with so much facility that it appears as if they are skimming and skipping around. It is poor readers who decode text using known words, context, pictures, familiar letter combinations, and plain old guessing. Their trouble does not lie in comprehending text; it is their inability to connect spoken with written language that frustrates them.

These children must be taught that individual sounds heard in spoken language—phonemes such as "ch"—can be represented by letter combinations, and that these sounds and letter clusters can be put together to form words. It is on this point that whole language and phonics-first teachers are most divided. Whole language instructs that phonics, if taught at all, should only be taught implicitly, allowing children to deduce the sound-symbol relationships through their engagement with text. The findings from the NIH directly contradict this. While many children easily grasp these connections, a significant number need them to be taught explicitly, says Lyon.

The Great Wall

So why the chasm between research and practice? How could a philosophy whose basic principles have been proven false survive and continue to gain supporters? Testifying before the Committee on Education and the Workforce in the House of Representatives this past summer, Richard Venezky of the University of Delaware said that part of the problem is that the government and various foundations fund the research but do not disseminate its findings. Indeed, during the committee’s hearing on literacy, chairman William Goodling noted, "I’ve been here all these years and never knew there was an ongoing project on reading at the NIH." Neither did his colleagues, and since only 13 members of the 45-member committee even bothered to show up to the hearing, few of them found out about it.

More disturbing than Congress’s ignorance is the situation on the front lines. The people who should be most familiar with the research—education professors, teachers, and school administrators—have routinely adopted instructional methods and curricula heavily influenced by whole language in spite of the overwhelming body of research evidence supporting phonics. It’s as if educators have erected the intellectual equivalent of China’s Great Wall, successfully thwarting researchers’ efforts to invade the schoolhouse.

In part this is due to a lack of leadership at the federal level. The government funds research at the NIH, at the Center for the Study of Reading at the University of Illinois (which hand-picked Adams to write Beginning To Read), and at the University of Oregon’s National Center to Improve the Tools of Educators, all of which have stressed the vital importance of early phonics instruction. Yet the U.S. Department of Education has hardly anything to say on the topic beyond reminding parents to read to their children. The result: Little of this research reaches the classroom.

State education agencies are equally hesitant to take a strong position. The Massachusetts State Department of Education refuses to make recommendations concerning instructional methods, and Patricia Webster of the New York State Department of Education said, "We don’t suggest from this level how reading should or should not be taught in the classroom. Decisions like that are left to the local districts." William Farr of the Connecticut State Department of Education labeled Reid Lyon, a division chief at one of the most respected research institutions in the nation, an "extremist," adding that the state department takes no official position.

Still, plenty of education professors are familiar with the research, yet fail to incorporate it into their classroom instruction. Which begs a question: Why is whole language so seductive? Researchers speculate that whole language’s popularity stems from teachers’ search for a method that’s easier for children and frees teachers from using the stuffy worksheets and dull drills of yesterday. "The whole-language movement should be about displacing compartmentalized instruction and rote facts and skills," write Adams and Bruck, phonics-first advocates. "And it should be about displacing such outmoded instructional regimens with highly integrated, meaningful, thoughtful, and self-gendering engagement with information and ideas." But Adams is quick to add that systematic phonics does not imply schooldays filled with painstaking recital of letters and syllables. All it takes, she says, is 20 to 30 minutes each day.

Whole language also flourishes because of the long-standing skepticism toward research in the education community. Even educational researchers admit to the shoddiness of educational research in the past, and the tendency of "the latest findings" to swing educators from fad to fad. "Very little research on anything ever makes it into the classroom. What you get are trends with very little research evidence to back them up," says Gerald Bracey, a research psychologist who writes widely on education issues. "Educators run from one fad to another," adds Smith of the NEA.

Further hampering attempts to extend research findings into practice is the fact that teachers are rarely taught how to read and analyze research evidence. For the most part, teachers colleges that serve as vocational schools are separate from research institutions, so professors who train teachers are insulated from professors who engage in research. "They’re different professions," says Adams, "They go to different conferences and read different journals. The people who are doing research work in education are not well-informed about the real problems and needs of schools and teachers." There are exceptions, such as the Columbia University Teachers College and the University of Virginia’s Curry School of Education, but the vast majority of teachers graduate from "normal schools" where training and research are not integrated. "Teachers see professors as very smart people who do very good work that has nothing to do with what they do," says Bracey.

With scant ability to discern research from opinion, teachers and even school superintendents often adopt unproven practices. "Unlike other research-based professions," writes researcher Bonnie Grossen, "our mechanisms for distinguishing fads that will probably fail from effective innovations are weak and ineffective." Hypotheses by education professors quickly become "theories," even though they have seldom been subjected to any rigorous testing. As the dominance of whole language shows, this has been particularly true in the field of reading. "Hard science is often alien to primary-level reading instruction, snake oil and charismatic solutions being preferred far more often than this country can afford," said Venezky in his congressional testimony.

At the graduate level, whole language continues to inform reading instruction. In fact, surveys of teachers and the textbooks they use in education schools confirm that most teachers are not taught systematic phonics and are hardly ever told that whole language instructional methods are even contested. Reid Lyon tells of his encounter with a California teacher seeking a doctorate in reading instruction who approached him after a lecture. Her face wet with tears, she told him that no one had ever exposed her to phonics-based instruction. "The majority of teachers we’ve talked to who have been trained over the last 10 years have never even discussed these issues," says Lyon. "Teachers are resentful that they haven’t been presented this in the past," says Louisa Cook Moats, the director of training at the Greenwood School, a teacher-training institute in Vermont.

Education schools are routinely criticized for their emphasis on theory over practice. This is especially true in reading. One survey published in the journal Teacher Education and Special Education (1989) found that less than 10 percent of teachers had ever seen their professors demonstrate methods of reading instruction tailored to children’s differing needs. Fewer than 5 percent said that what they had learned about teaching reading actually related to what they did in the classroom. Remedial and Special Education published a survey in 1992 of 100 learning-disability experts, including many professors of reading instruction. It found that few assigned any importance to understanding basic language structures like syllables and phonemes. "The most common comment I get is that nobody ever taught me any of the substantive part of what it means to teach students to read," says Moats, who is directing NIH’s study of reading in the District of Columbia’s public-school system. "They continually ask, ‘Why didn’t anyone teach me these things?’ "

Good question. Part of the answer is that states rarely require more than one semester of reading instruction to obtain certification to teach. Another part is that few school districts evaluate teachers using student performance as a measure. In short, professors are wholly unaccountable for the teachers they graduate. They thus have little incentive to indulge in potentially tedious practice sessions and lessons in how to perform an effective phonics drill. "The average person who’s teaching reading on a university faculty knows very little about linguistics," says Venezky, now serving a year-long post as a resident scholar at the U.S. Department of Education. "To them, phonics is very often frightening, foreign, and very difficult to teach." Evidently, education professors cannot imagine themselves holding up that hungry bear named Chuck.

Phonics in Action

In January, Secretary of Education Richard Riley traveled to Houston, Texas, to laud the first city to accept President Clinton’s reading challenge. Curiously absent from the festivities was the Houston education system’s shining star, former Wesley Elementary Principal Thaddeus Lott. During the early 1980s, his success in turning Wesley from a typical urban failure into one of Texas’s highest-performing elementary schools led almost 300 Houston schools to follow his lead in abandoning the school district’s recommended curriculum.

Administrators instead used their discretionary funds to purchase DISTAR (Direct Instructional System for Teaching and Remediation), the program Lott had introduced at Wesley in 1975. DISTAR, now known as Reading Mastery, is a direct-instruction program developed in the 1960s by Siegfried Englemann, a former preschool teacher. It incorporates intense, systematic phonics instruction into a fast-paced, heavily scripted program with constant teacher-student interaction.

Lott’s success was swift. In 1980, just three years after the school’s third graders were first taught using DISTAR, 85 percent passed the reading comprehension portion of the Texas Assessment of Academic Skills (TAAS), up from 18 percent in 1977. In 1996, 100 percent of Wesley’s third graders passed the TAAS, even though more than 80 percent of the kids in the district are poor enough to qualify for free or reduced-price lunches. "Direct instruction has filled the void left by colleges and universities and teacher training institutions," says Lott, who is now franchising success by running Wesley as one of four charter schools under his control. "We’ve been fighting this whole language wave for years. People who only know how to teach whole language don’t know how to teach phonics."

Observers tend to chalk Wesley’s excellence up to Lott’s charisma and sense of mission. But his results should not be so surprising; Lott simply adopted the best program available. In 1977, Project Follow Through released the results of its decade-long, $500-million study of teaching methods that began as part of Lyndon Johnson’s "War on Poverty." The federal government study rated direct instruction the best method by which to improve student performance. Literature-based programs that were avowedly child-centered rated lowest. Paradoxically, direct-instruction programs produced the greatest improvement in student self-esteem, while the child-centered methods that claimed to raise student self-esteem ranked much lower.

Still, direct instruction has suffered from criticism that its strict program handcuffs teachers, and few schools have adopted it nationwide. Here’s the typical response from a direct instruction teacher, though: "The bottom line is that when I get third graders reading on a first-grade level, I don’t have time for flexibility," says Dianne Bissell, a former elementary-school teacher at one of the schools in Houston that adopted DISTAR after observing Wesley’s dramatic improvement. "Direct instruction works."

These schools purchase DISTAR with Title I funds granted by the federal government to provide educational opportunities to low-income children. At other schools, most Title I money is wastefully spent on early intervention programs such as Reading Recovery (which is estimated to cost between $8,000 and $9,000 per child), teacher aides, and remedial help long after intervention would have been most effective. The NIH has developed tests that cost $15 per student to assess whether children will have trouble learning to read in first grade. When used in conjunction with an early intervention program such as DISTAR, Lyon says, 85 to 90 percent of poor readers can reach average levels if diagnosed early enough. As time goes by, the costs rise as the probability of success declines.

Bridging the Chasm

The "reading wars" have become a political battle rife with smears and misrepresentation. Whole language was blamed entirely for sinking reading scores in California, even though the state also failed to train teachers in the new literature-based curriculum. Critics of the public-school system such as Samuel Blumenfeld view whole language as simply the latest attempt by the education establishment to "dumb down" America’s children, clearing the path for a socialist revolution. Whole-language adherents tend to disclaim the validity of any scientific study and accuse phonics-first supporters of a hidden agenda: delegitimizing the public schools to win funding for private religious schools.

But basic skills are an issue of common sense, not conservative policy. Consider the sport of wrestling. A wrestling match between two skilled athletes may appear to a casual observer to be as natural as a good reader breezing through text. What the observer does not know is that a wrestling match is made up of dozens of individual moves and skills that the competitors have practiced for much of their lives. Wrestling coaches teach all of these skills in isolation and then let their wrestlers practice them on the mat. After hours of practice and drill, these skills are so automatic and fluid that the wrestlers do not even need to think about them during a match. They can worry about broader strategy—just as a reader with excellent decoding skills can concentrate on comprehension. From sports to driving to chess, in no other field except reading would teachers tell their students that learning the basic skills first is unimportant, perhaps even harmful.

It’s true that basic skills are no panacea. Reading is a complex activity, and a host of other factors impact how well children learn the relationship between their spoken language and its representation in print. Most important is a child’s readiness upon entering school. A 1995 study found that children in professional families heard 2,150 words per hour on average, working-class kids were exposed to 1,250, and children on welfare heard only 620 words per hour. Upper-income parents were also much more likely to ask stimulating questions and challenge their children’s cognitive skills. These kids stepped through the kindergarten doors far more experienced with language, giving them a tremendous advantage in the acquisition of reading skills.

While bearing in mind the effects of the home environment on learning, it is equally important to note the further damage caused by receiving ineffective methods of reading instruction. The NIH studies have proven that poor children whose parents do not expose them to books and language suffer the most under programs like whole language that do not emphasize skills. Their difficulties are compounded when they reach higher grades and have yet to learn the fundamentals of reading, hindering their study in all other subjects. And Thaddeus Lott’s success with direct instruction proves that a disadvantaged background does not prevent children from succeeding alongside their suburban peers—if they are taught using research-based methods.

Egalitarians worried about the increasing distance between rich and poor should take heed of researchers’ warnings. Current methods of reading instruction are exacerbating differences in educational opportunities, allowing the well-heeled sons and daughters of loquacious professional parents to reap the advantages of wealth while impoverished children linger behind.

As Flesch wrote more than 40 years ago, "There is a connection between phonics and democracy—a fundamental connection. Equal opportunity for all is one of the inalienable rights, and the word method interferes with that right. . . . [I]t returns to the upper middle class the privileges that public education was supposed to distribute evenly among the people."