The Reading Wars: Understanding the Debate Over How Best to Teach Children to Read

Los Angeles Times

By Kenneth Anderson
Los Angeles Times
Sunday, June 18, 2000
Kenneth Anderson teaches law at American University, Washington D.C.

What is it about teaching reading that arouses such passions in Americans? Shall we have phonics or whole language or both? Why this debate should be so vehement in the political arena is not immediately obvious. Nor is it obvious why the issue is so important that George W. Bush, for example, has been running television ads prominently featuring phonics, as though it were a topic as central to the presidency as social security, taxation, trade with China or nuclear weapons. One answer is simply that Americans are anxious about the primary skill necessary for their children’s success. This anxiety requires a dispassionate answer to the critical question: How can schools best teach reading?

More than 15 years worth of books, articles and learned studies have sought to declare peace in the reading wars between various instructional techniques, most notably phonics (which teaches word-decoding skills before textual meaning) and whole language (which emphasizes textual meaning). No peace is forthcoming, however. The reading wars, it is clear, are waged not on the front of technique alone but as part of the culture wars in which traditional versus progressive pedagogy, and the values they embody, are at odds in America’s schools.

In the political arena, it would be fair to say that phonics is winning. California is the leading edge of the movement, but the rest of the country is not far behind. Implementation is, however, another story entirely. Much of the educational establishment has fought, and continues to fight, a rear-guard action to preserve whole language pedagogy, mostly by changing the names of programs without changing their substance and partly out of a commitment to the ideal of progressive education.

Though any generalization does a disservice to the complexity of the issues, those who favor traditional values in education tend to favor phonics instruction, and those who favor progressive, or child-centered, values in education tend to favor whole language. But it is also true that many of those who favor phonics on the basis of its effectiveness as a teaching tool—especially some research scientists—are relatively indifferent to the larger values at issue and may have no sympathy with the idea of “traditional” education at all. Nor do they necessarily have any desire to be associated with the religious conservatives who strongly supported the back-to-phonics movement more than a decade ago.

The movement for a return to phonics simmered for many years among those philosophically opposed to what they perceived as the lack of discipline in whole language pedagogy; as early as the 1970s, too, a few academics were growing skeptical of whole language claims of reading outcomes and achievement. The movement gained momentum in mainstream America in the last 15 years, however, especially among parents and especially in California, where whole language was widely perceived by the public to have been responsible for a sharp downturn in reading scores even among middle-class, suburban, native English-speaking students. A causal relationship between whole language and reading score declines was insisted on by some researchers and sharply disputed by others, but by the late 1990s, legislators had responded in California and many other states with mandates to teach or emphasize phonics.

Winners or losers notwithstanding, the passions will not go away, and Maureen Stout’s “The Feel-Good Curriculum,” a scathing polemical survey of where American education stands today on the question of which values are central to the education of children, helps us understand why. Hers is a story that addresses the values of progressive education in America’s public schools that have been championed during the last 100 years. Stout chronicles the triumph of the “therapeutic” in American schools, a trend that focuses more on how the child feels than on what he or she has learned, which, during the last 20 years has crystallized as the self-esteem movement in schools. The ideological centerpiece of that movement, as Stout documents with great thoroughness and sharp criticism, is that self-esteem precedes achievement and is something that can and ought to be instilled independent of achievement.

Much ink has been spilled attacking and defending this view, especially in California. Though the self-esteem movement has lost much of its initial public acceptance and become something of a joke, it remains institutionalized in the schools. Its central assumptions lie at the center of the reading wars. Whole language pedagogy incarnates, within one specific part of the curriculum, the self-esteem model and progressive education. Stout, drawing on John Dewey’s classic formulation in “Experience and Education” (1938), writes that it is characterized by the “cultivation of individuality, free activity as opposed to external discipline, learning from experience rather than from texts and teachers, acquiring skills that are deemed relevant to the individual at the present time rather than preparing for some unknown future, and becoming acquainted with the world rather than learning through static aims and old materials.”

Within the dominant structure of progressive education, phonics is regarded as the polar opposite of whole language; it is rigid, authoritarian and fanatically concerned with the acquisition of skills such as spelling. Phonics is seen as deeply anti-democratic, and its critics, defenders of whole language, find it inconsistent with the abstract values of progressive education.

Indeed, one can detect such tones, for example, in a recent Washington Post editorial offering guarded congratulations to the home-schooled, religiously conservative children who took several top places in both the national spelling bee and national geography competition a few weeks ago. Of course these children excelled at spelling words such as “apotropaic,” the editorial seemed to say, because rote training, mindless memorization and obedience to authority are what their traditional education (including, naturally, phonics) is good for.

Progressively schooled children may not spell well, but they do have democratic, multicultural values. Or as Stout correctly observes: “Until very recently anyone who ventured to suggest that phonics is still the best foundation for teaching reading was regarded as ill-informed or worse—uncaring.”

But the nature of the debate is changing, partly as phonics advocates have won the political wars and partly as scientists, who are understanding how written language is actually acquired, have de-coupled the “effectiveness” debate from the “values” debate. Today the debate over reading pedagogy has fractured into three main strands. First, there are those seeking to split the differences between methodologies and declare peace in the reading wars by accepting that all reading methods are important and should coexist in the classroom. Second, there are whole language advocates, down but not out, alternating between strategies of defending whole language on the grounds of progressive education, “caringness” and democratic values, on the one hand, and attacking the scientific claims of phonics-oriented researchers on the other. Third, there are phonics advocates, who understand that they have politically and intellectually won the war with whole language supporters but who must now confront the difficult and uncertain task of finding the best phonics pedagogy to consolidate their victory.

Ultimately, the most important and exciting debate in the reading wars today is not between phonics and whole language but instead within the phonics movement, as it struggles to create practices that will make phonics a tool of reading success rather than simply another forlorn experiment in American education.

Nowhere is the anxiety over reading pedagogy more evident, despite an attempt to bury it beneath the surface, than in the long-awaited report of the National Reading Panel, “Teaching Children to Read,” released just a few weeks ago.

Established by Congress in 1997 to undertake a comprehensive review of research on how best to teach reading to children in America, the panel, organized under the auspices of the National Institute of Child Health and Human Development of the National Institutes of Health, consisted of 14 individuals drawn from leading scientists in reading research, representatives of colleges of education, reading teachers, educational administrators and parents. The resulting review of the research literature is the most comprehensive and rigorous to date—rigorous in no small part because it accepted only studies using standard social science quantitative design criteria.

Notwithstanding its methodological rigor, the panel was structured to give a place to all viewpoints and political constituencies in the perennial debate over reading pedagogy. The panel divided its review work into topics. The most important and politically fraught, not surprisingly, was the argument between phonics and whole language. Other topics included newer questions, such as the training of teachers and the use of technology in teaching reading.

The result of the political structure of the panel—despite its close association with the National Institute of Child Health and Human Development, a long-time funder of phonics-favorable research—is a report that is anodyne, bureaucratic, soothing and cautious. Product of committee and consensus decision-making, the report attempts to pour oil on troubled pedagogical waters but is largely unable to reach conclusions that are plain and robust enough to serve as foundations for public policy. Everything is endorsed and nothing is rejected, an “I’m OK—You’re OK” report on reading pedagogy. Phonics is deemed critically important, as is the ability to apply whole language strategies to “enhance understanding and enjoyment of what children read.” The report gives the impression of having been written one paragraph by one political faction, the next by another, with each sentence negotiated phrase by phrase; a corporate lawyer would instantly recognize the genre. Congress might have done better to have asked the panel to determine which methods of reading do not work or which work less well than others, in order to force a priority of methodologies.

Nevertheless, amid the decorous congratulations offered to virtually all reading methods, there is a faint but crucial ordering of the pedagogies. Phonemic awareness skills (the ability to “manipulate the sounds that make up spoken language”) and phonics skills (understanding that there are “relationships between letters and sounds”) are given a marginally greater endorsement than other methods. Phonemic awareness is deemed “essential”; the research review showed that “teaching phonemic awareness to children significantly improves their reading more than instruction that lacks any attention to it .”

Phonics strategies for reading must instill two skills: one, the ability to decode a word on the page and the ability to say a word and then spell it. In English, this is particularly difficult. The sound of “read,” for example, is pronounced with a long E or short E, depending purely on the intended tense, and no amount of phonics rules will remedy that. The spelling of homophones (words that sound alike but have different meanings such as “wait” and “weight”) depends ultimately on their usage. Given the orthographic ambiguity of written English and phonics’ inability to sort out when “too” is “too” and not “two,” it is no wonder that educators threw up their hands and turned to a whole language approach, believing that the promise of consistent mapping rules was a cruel hoax on children. Yet it does appear that contemporary phonics systems radically reduce the indeterminacy of the written language.

It is one thing to talk about reading and another to talk about teaching reading. Reading has no point if it is merely decoding without comprehension. The issue is whether the priority in teaching reading, in the initial stage of formal reading instruction, especially in the first grade, should be the teaching of phonemic awareness and phonics skills that, in the view of phonics instruction advocates, make reading and then reading with comprehension possible in the first place. This unavoidably pits individual word-decoding skills against the comprehension of a whole text, at least in the initial stages of instruction. It is a competition that book after book, seeking to declare peace in the reading wars, tries to make go away but without success, because although it is a truism that reading requires both, the teaching of reading inevitably begins by giving one priority over the other.

On this issue, the report does the public a great disservice, first by pretending such prioritizing does not exist and next by wrapping itself in contradictory platitudes about the benefits of “systematic” phonics instruction while simultaneously warning against making systematic phonics, for example, the “dominant” force in the first-grade classroom. Moreover, the struggle between phonics and whole language cannot be simply resolved, as some people might wish, by announcing that they can coexist within a classroom. This norm, seen as the best of both worlds which schools are moving to adopt as a means of responding to political pressure for phonics, is a mistake. Phonics and whole language proceed from fundamentally different assumptions, and each claims priority on the attentions of the first-grade teacher and reader. “Mixed” is a recipe for the worst of both worlds, not the best; one pedagogy is inevitably primary and the other secondary in the classroom, and it matters profoundly which one chooses as the priority for a particular grade and age. Any reading program necessarily does choose, and we would be better off if the report had said so.

The one clear conclusion that the report reaches—and it is not insignificant—is that without some phonics instruction, whole language pedagogy is not enough. Indeed, the report finds the characteristics of phonemic awareness training “most effective in enhancing” reading and spelling skills, “included explicitly and systematically teaching children to manipulate phonemes.” Italics added.

Having reached these conclusions, however, the report promptly qualifies them. It cautions that intensive, systematic phonics—the kind that parents are wont to demand—may not be a good idea. In particular, according to the report, allowing phonics to become the “dominant component” of reading instruction, particularly in the first grade, may be an especially bad idea if it is at the expense of reading activities that focus, for instance, on meaning.

After praising the systematic and explicit teaching of phonemic awareness, the report then endorses exactly its opposite: “embedded phonics instruction,” whole language pedagogy in which phonics is supposedly taught but only “implicitly,” not systematically and, many observers would say, in reality not at all. Everything, seemingly, is a pretty good method of teaching reading if it has a powerful constituency behind it.

Still, press coverage of the report showed that the media were able to grasp its modest preference for phonics. Education Week, for example, headlined its story “Reading Panel Urges Phonics For All in K-6,” and numerous editorial pages around the country cited the report as an endorsement of systematic phonics to the irritation of leading whole language advocates, such as Cathy Roller of the International Reading Assn., who correctly pointed out the many passages in the report that favored whole language. One member of the National Reading Panel, an elementary school principal, felt compelled to issue a minority report essentially criticizing the panel for taking the side of phonics over whole language.

Whole language pedagogy has profound problems, however, that not even the soothing language of the report can make go away, not least of which is that the theories of written language acquisition on which it is premised appear to be wrong. Whole language theories of how children learn to read and write begin with the developmental assumption that just as children are neurologically wired for spoken language, they are similarly hard-wired for reading and writing. Reading, then, is a natural process that ordinary children will painlessly acquire without explicit instruction, without systematic training but instead, for example, by reading aloud and being read to, so-called “guided reading.” Advocates of whole language emphasize that whole language makes reading inviting and fun because it begins with real stories and real characters. It draws beginning readers into the joy of reading, rather than numbing them into inattention and unmotivated boredom with the excruciatingly artificial non-stories of the phonics primers of the past. Whole language is able to favor the fun and joy of reading from the very beginning, it is argued, because it starts from the assumption that learning to read and write, being natural proclivities of children, requires no arduous intellectual discipline.

Despite all the public debate and parental clamor for phonics instruction in America’s schools, whole language remains dogma in the nation’s graduate education schools; it is what the educational establishment in America still wants to believe is the most effective strategy. At the first-grade orientation meeting just a year ago at my daughter’s fancy, trendy private school, I listened in astonishment as the first-grade teachers—many of them recent, top-ranked graduates of some of the country’s most prestigious schools of education—informed parents that reading is a natural developmental process that would unfold all on its own, in its own way and time. Children would teach themselves to read, without explicit or systematic instruction, in a process as natural, we were told, as “teeth coming in.”

I am not aware of any neurolinguist who holds this view. I doubt any have held it in decades. The consensus among scientists is closer to what the well-known cognitive neuroscientist Steven Pinker has written: “Language is a human instinct, but written language is not. Language is found in all societies, present and past. . . . All healthy children master their own language without lessons or corrections. When children are thrown together without a usable language, they invent one of their own. Compare all this with writing. Writing systems have been invented a small number of times in history. . . . Until recently, most children never learned to read or write; even with today’s universal education, many children struggle and fail. A group of children is no more likely to invent an alphabet than it is to invent the internal combustion engine. Children are wired for sound, but print is an optional accessory that must be painstakingly bolted on. This basic fact about human nature should be the starting point for any discussion of how to teach our children to read and write.”

One wonders what the report would have said had it been written not by a mixed group of educators, teachers, administrators, parents and scientists but instead by a group of neurolinguists. Would it have caviled quite so much on the question of whether phonics or whole language is the correct approach, or whether phonics instruction should be systematic and explicit?

It does seem likely, however, that a pedagogy based on hard science would begin from assumptions far different than those of whole language. It has no doubt been disconcerting for the advocates of whole language instruction to find that over the last decade, the better part of expert and scientific opinion has moved against them.

Partisans of whole language pedagogy, dedicated educators to be sure, have been fighting a rear-guard battle, based upon the view that children are not developmentally equipped for the discipline required by phonics instruction in the first grade and that the far greater risk to children’s ultimate ability to read is—in a world in which reading must compete with television, video games and a thousand other entertainments—boredom and the inability to see the point of phonics drills. Typically there is resistance to the mandates to teach phonics issued by state legislatures and state boards of education. Though California is a leader among states in the revival of phonics, the effective consolidation of phonics is still a long way off and requires overcoming both the genuine difficulties in promulgating a new and demanding curriculum and the passive resistance of many educators. The easiest path is to teach whole language and call it phonics. Meanwhile some voices have emerged to defend whole language pedagogy publicly and unapologetically; one of the most well-known is the educational psychologist Gerald Coles.

“Defender of whole language pedagogy” is not a characterization Coles would always have accepted. In a 1998 book, “Reading Lessons,” Coles rejected as inadequate the battle between phonics and whole language. Instead, he argued, the real issues in teaching reading transcend the mere pedagogy used in the classroom and are best understood in terms of race, equality and inequality of income and political economy and political power. As a consequence, “Reading Lessons” gives brief histories of the phonics and whole language pedagogy only to dismiss them both in favor of a broad discussion about poverty and money in education. As “Reading Lessons” says, drawing on the work of Soviet psychologist Lev Vygotsky and his followers:

” The conception of ‘interactivity’ emphasizes the importance of societal organization and power for classroom teaching and learning, even when the influences of that organization and power are not readily apparent. This does not mean that societal influences determine smaller units of teaching and learning literacy. Rather, they contribute mightily to, and therefore are inseparable from, the interactivity that comprises literacy teaching and learning.”
There is something right about this. It is genuinely important to understand the interconnections of wealth, opportunity, race and class in society on the one hand and their reinforcing interactions with such micro-phenomena as teaching reading on the other. Although I would analyze these factors differently than Coles, I am mindful that my daughter’s school—the very embodiment of whole language and progressive education—does not do a very good job of teaching reading. Yet, by and large, the parents do not much care. They do not care precisely because of the issue of “interactivity,” which includes their ability to provide private tutoring and other mechanisms to overcome any deficiency of the school. Income inequality enables rich families to remedy inadequate instruction. The poor have no such option, however; economic disparity prevents less advantaged parents from buying themselves a new reading pedagogy.
Coles’ argument is actually an argument for school choice, vouchers and other funds “strapped on” to students so that even the poor might have some part of the choices available to the rich, although this is something the progressive Coles would recoil from. Even without driving toward conclusions he would find politically untoward, however, Coles’ arguments run the risk of promoting quiescence and passivity: If improving the state of reading education requires massive social changes, then why try to do anything as minor as seeking the best pedagogy for teaching reading in individual classrooms and schools?

Perhaps sensing the force of such objections, Coles has retreated to the robust defense of whole language pedagogy. In “Reading Lessons,” he argued that the very debate between phonics and whole language was sterile; today he has become an intellectual spokesman for whole language. His new book, “Misreading Reading,” is a crisp salvo fired across the bow of neurolinguists and other scientists who back phonics instruction and the research sponsored by the National Institute of Child Health and Human Development. The research is, in his view, very bad science, and his “Misreading Reading” is, in essence, a handbook for whole language educators looking for materials with which to debate parents and others calling for phonics instruction. It has summaries, charts and diagrams which, the publisher points out, can “serve many possible uses—including ‘talking points,’ handouts or overheads for presentations.”

But in attacking science that supports phonics, Coles comes perilously close to endorsing the whole language myth that reading and writing are “natural.”

“Whole language maintains that children’s motivation for learning written language is similar to that which impelled their learning of oral language: the desire to make meaning in order to participate and communicate within a community of language users. Just as their making of and communicating with oral meaning was the overarching orchestration that promoted their pronunciation of words, use of words to identify multiple objects, stringing words into sentences, syntax, and vocabulary development, and so forth, making of and communicating with written meaning has similar effects.”

It is hard to imagine Pinker agreeing. Coles has fatally conflated what Pinker called the “language instinct” in humans for oral language with a “desire” to learn how to read and write. They could not be more different. One is a biological imperative for members of our species; the other a contingent social construct that has appeared only fitfully in our history, socially created and inculcated into children with great effort and frequent failure.

Still, Coles is content most of the time to reconceive whole language pedagogy as the modest proposition that students ought to understand what they read. He tells us that whole language has suffered from numerous misconceptions in the public mind, including the “belief that whole language teachers simply create a print-rich environment and then let children intuitively learn to read.” On the contrary, Coles says, whole language pedagogy—citing its advocate Regie Routman—”teaches the ‘basics’, but in meaningful literary contexts; provides explicit instruction when students’ needs and interests require it; provides rigorous teaching with high expectations; and is greatly concerned with learning outcomes.”

In the last few years, to be sure, the definition of whole language has changed to fit today’s changed political environment. Perhaps what Coles writes is now true; under intense political pressure whole language has rehabilitated itself as “phonics with a happy face,” an non-controversial pedagogy which accepts direct instruction, the “basics” and even the notion that children ought eventually to learn to read books without having to resort to illustrations to provide context. Who could possibly object to this reformed version of whole language—the prettified version that figures throughout the report—now shorn of any theoretical pretension other than “understanding is good”?

But many people would disagree with him. Many would say that the practice of whole language continues to eschew both the “basics” and explicit instruction, still provides little teaching other than “guided reading,” promotes by default the memorization of whole words in lieu of learning how to decode individual words and writes off the many children who do not learn to read under whole language methodology as having learning disabilities.

But ultimately, a rehabilitated theory of whole language instruction is irrelevant. The only real issue for whole language (or any other pedagogy) is whether it should have pride of place in a national curriculum or—even more radically—be the sole pedagogical method for teaching reading without having to assert a theory of how written language is acquired.

Phonics has such a theory. Whole language had one for decades, but it is now intellectually discredited, and Coles is among the few sophisticated writers willing to defend even a modest version of it. But if all that survives of whole language pedagogy is that children should learn to comprehend what they read, then whole language loses any distinctive pedagogical claim and becomes secondary to the method which has an explicit and scientifically plausible account of how children acquire written language. After all, what made whole language intellectually distinctive was not only its “reading is natural” assumption but also its argument that phonics, formal rules for decoding words, do not have to be taught. Whole language is a pedagogical theory that excommunicates phonics as unnecessary.

Of course, to the extent consistent with the priority of learning decoding skills, phonemic awareness and phonics skills should be taught in a context that is interesting and stimulating, with real literature that invites understanding. Children who do not come from print-rich and literate environments, who have no reason to think that reading is important to them and whose parents do not read to them need the invitation of exciting, imaginative literature to give them a reason to do the harder work of phonics instruction. Particularly among children from deprived homes, the classroom will have to make explicit the connection between the discipline of phonics and its eventual results in the joy and necessity of reading that other children will discover at home.

One of the peculiarities of whole language pedagogy is that it takes informal reading, which committed parents have always done with their children, and puts it at the center of the classroom experience, while abandoning (to parents, if to anyone at all) the systematic and explicit instruction in phonics that the classroom is far better able to deliver than parents. A justification for this curious inversion has been that because so many parents no longer undertake reading activities, preferring to fetishize the television rather than the book, the school must take over that function too. This is an unfortunate trend but not a reason for schools to drop what they can do best.

The rest of “Misreading Reading” is a detailed discussion of the scientific studies on phonics. Coles is not impressed; he argues that it is bad science that has been misused in public policy, pointing out that there is a history of phonics-oriented researchers overselling their results, with the media leveraging the oversell even higher. He is right to emphasize, as he did in “Reading Lessons,” that no mere reading pedagogy will overcome the effects of poverty and deprivation in the reading lives of poor children. Yet though he has read widely and pondered details, he is finally not persuasive. He is useful and talented as a critic but does not extend his critical review of the literature to the findings made by researchers favoring whole language, which are even more extravagant in their conclusions and much weaker in their methodologies than phonics-favoring research.

Coles seems to think that weaknesses and gaps in the scientific studies backing phonics instruction ought to cause a retreat to whole language; this does not follow. Still, he makes an important point that within the phonics movement, there is wide variance in theories, teaching methods and instructional materials and that not all of them are likely to be good, let alone the best. Leaving aside those who will continue to teach whole language while calling it embedded phonics, there are many serious questions about how best to teach phonics. There is no agreement on how to systematize, for purposes of first-grade instruction especially, the components of phonemic awareness and phonics skills. Phonics has won the intellectual debate with whole language, but that does not mean that the best method of phonics instruction is clear.

Nor is it clear that the way phonics was taught in the past is the best way to teach it today. Conservative educators ought to be cautious in announcing, as Diane Ravitch once did, that success in reading means abandoning the failed methods of whole language and using instead “time-tested instructional methods in reading.” Research is more likely to point toward instructional methods that are unambiguously phonics-centered yet differ significantly from those of the past. Schools will not be saved by resurrecting the 19th century McGuffey Readers.

The stakes in this search for method are enormously high. Phonics instruction works best as one-on-one instruction by teachers, but there are few teachers who know how to teach phonics using any established system. The shift to phonics requires retraining a whole generation of teachers without clear maps to guide them. Critics such as Coles await, and teachers unions, never greatly interested in direct instruction, will be eager to pronounce failure and call for a return to less strenuous pedagogy.

Given the uncertainties, perhaps the most important thing that can be said today is that phonics instruction should be systematic, explicit and—contrary to the report—intensive. A great many books offering systems for teaching phonics have appeared in recent years, and they vary greatly. One that is comprehensively argued but still pitched to parents and non-technically trained educators is Diane McGuinness’ “Why Our Children Can’t Read,” with a foreword by Pinker expressing his views on the differences between spoken and written language acquisition by children. Its aims are not modest.

The first section examines writing systems across a wide variety of cultures and societies to show why “children easily become confused about our writing system and fail to learn to read and spell.” It provides an analysis of what McGuinness calls the “alphabet code,” how the sounds of the English language are “mapped” to its alphabet. It examines the scientific literature and relays how a child might effectively learn to deploy the alphabet code in reading and spelling, skills which McGuinness concludes “are highly trainable” with systematic instruction. Finally, it applies these findings to pedagogy and sets out both a model program for classroom or home use and for remedial work with children who have difficulty reading. The final chapter addresses itself to parents seeking to either avoid reading problems altogether or remedy ones that exist.

“Why Our Children Can’t Read” is not a book mired in the past. One of its most interesting parts is an analysis of what’s wrong with Noah Webster’s hoary 18th century “The American Speller,” which has provided the template for numerous contemporary phonics methods. McGuinness’ discussion of Webster’s errors provides an insight into what she thinks is critical to phonemic instruction: “The Speller itself consisted of word lists of increasing complexity and syllable length. . . . Rather than organizing the lists by sounds (phonemes) in the language, Webster decided to do this by letters, word families, and syllable lengths, set out alphabetically. In some ways this is understandable, because it is hard to represent ‘sounds’ in print. . . . What happened as a result of Webster’s emphasis was that the sounds of the language . . . got lost.”

As a result, according to McGuinness, the problem with traditional phonics classrooms is that the “alphabet is taught entirely from letter to sound, which destroys the logic of the alphabet code. . . . There are diametrically opposing ways to teach ‘sounds.’ One way, traditional ‘phonics,’ is to teach ‘the sounds of letters.’ The other way, which doesn’t yet have a name, teaches the ‘sounds of the language’ and how those sounds are mapped to letters. Misunderstanding this fundamental difference has been the major cause of failure of reading programs probably well before Webster. In this century, it has been the cause of the endless flip-flops between whole-word methods and phonics.”

McGuinness’ critique of traditional methods of teaching phonics—that they invert the logic of mapping the sounds of a language to its letters—is a powerful way to understand the limitations of the many phonics instruction systems which do not start from the sounds of the language. One persistent error, in McGuinness’ view—and a central feature of many programs aspiring to be pure phonics as well as many unsystematic phonics-whole language programs (and one phonics method endorsed by the report)—is teaching “analogies” or “word families” or, more technically, “rimes.” This system (“The Cat in the Hat” was based on one) often teaches rhyming words. For instance, if the child can learn to recognize and sound out the letters UNCH, then the words “bunch,” “hunch” and “lunch” can be more easily deciphered. The problem with this very common way of teaching phonics, McGuinness says, is that “children are taught, or led to believe, that these letter sequences are only one sound.” This means that if they are taught UNCH, they may have difficulty understanding similar sounding endings like ENCH (“bench”), ANCH (“ranch”) and even AUNCH (“launch”) because they learned UNCH as one unit or one “sound.”

” E ach word family or rime has to be taught one at a time,” McGuinness writes, “but there are 1,260 possible rhyming endings in English. Even if teaching rimes was an efficient way to teach spelling patterns (which it isn’t), phonics programs never teach more than a fraction of them.”

Her point, of course, is that UNCH is not one sound. Fully decoded, it has three; leaving aside technically accurate representations of the sounds, they are variations of U, N and CH. Full comprehension of the alphabet code requires training children to hear and decode those smaller units, and mere rhyming cannot do this.

Another popular strategy in current phonics programs emphasizes that “longer words can be broken up into smaller words or word fragments.” Having discovered various letter patterns, a child may then attempt to read longer words by breaking up words into one or another smaller words or fragments, hoping by trial and error that one of them will sound familiar and so indicate the word being read. The child who undertakes this phonics strategy is described as a “part word assembler,” and McGuinness observes that when he “reads ‘hated’, he sees ‘hat’, ‘ate’ and ‘ted’, and so he reads ‘hat-ate-ted.’ He also knows he isn’t reading real words. He is hoping that what he reads will sound similar to a real word so that he can figure it out. He tries different options.”

Part word assembly is a phonics strategy, to be sure—just not a very good one. It depends crucially on guessing, and sometimes it works; sometimes it doesn’t.

McGuinness’ candid and extensive criticisms of existing phonics programs are her most valuable contribution. The case that McGuinness makes for starting with the sounds of the language rather than the older alphabet-based systems, including Webster’s and his numerous inheritors, is persuasive and increasingly accepted. One sees a growing number of instructional systems broadly sharing this approach. But, as she points out, this is not enough by itself. A successful system must teach children the smallest units of sound in the language, the phonemes themselves. Anything short of that—as the examples above show—will be phonics and yet will ultimately be unsuccessful in teaching reading and spelling.

McGuinness believes that starting by teaching language sounds, and not alphabet systems, carries a further profound consequence for phonics. In her view, phonics programs should start with those smallest sounds of language; starting instead with combinations of those sounds—rimes or syllables, for example—will undercut a child’s ability ultimately to learn the phonemes themselves. McGuinness is adamant that ‘mixed’ classroom methodologies must be avoided. They encourage children to memorize whole words, a practice that gets more difficult, if not impossible, as more and more words are taught. She insists children must be steered away from this practice if they are to learn how to decode words using phonemes.

“Children are unaware of phonemes in speech, and it is easier for them to become aware of syllables or whole words. If a child can only hear words or syllables, he or she won’t understand how to use our writing system. For this reason, no reading method should ever teach children to read whole words, syllables, or syllable parts like ‘rimes.’ These are the wrong sensory units for our writing system. . . . Children must be trained from the start to become aware of the individual phonemes in speech. The earlier this is done, the easier it will be for a child to learn to read.”

“Why Our Children Can’t Read” is not without its controversies and weaknesses. McGuinness overstates matters, for example, when she argues that dyslexia is not a “causal” diagnosis but simply a description. She says, repeatedly, that what is called dyslexia is nothing more than bad instructional methods and that teaching the alphabet code would cause the problem to disappear in toto. The hard science on that issue does not permit a firm conclusion, as Coles correctly points out in “Misreading Reading,” although McGuinness is surely right that it has become the label of choice when whole language programs fail children. An honest observer must admit, however, that dyslexia will regrettably become the explanation when one system or another of phonics, whether badly designed or badly implemented, fails children. Faced with a diagnosis of dyslexia or, more generally, a learning disability, too many parents fail to ask the school an important question: “Well, what, exactly, did you teach my child?” (Parents who can afford remedial tutoring would be well-advised, before accepting a diagnosis of dyslexia in an otherwise normal child, to read McGuinness’ chapters on remediation and work with a competent tutor using a method that systematically teaches the full alphabet code, from sound to alphabet, building up from the level of individual phoneme.)

The risks of today’s passion for phonics instruction are threefold. First, as McGuinness demonstrates, phonics as currently disseminated is by no means as effective as it could be and, in some cases, is probably not effective at all. Second, the phonics movement risks promising more than it can realistically hope to deliver, especially to poor children. (Phonics systems such as Open Court, for example, have delivered promising results thus far in disadvantaged school districts, such as the Inglewood Unified School District in L.A. County, but it is much too early to extrapolate anything but hope from those preliminary samples. An equally important test of phonics as a curriculum will come in California’s suburban middle-class schools, where the slide in reading scores has also been serious.) Third, critics such as Coles are right to say that no mere curricular reform by itself can compensate for the effects of deprivation on poor children. Phonics risks selling itself as a panacea for poverty.

Reading is both an acquired taste and an acquired skill. Taste is fed by skill, and skill requires discipline in acquisition and discipline in teaching. Discipline is at issue in the reading wars; those who insist on seeing the reading debate as more than merely a matter of the most effective pedagogy and instead as something as personal as one’s values, are not wrong. One pedagogy requires intellectual discipline as a condition for learning, while the other denies that this discipline is necessary and sees it as harmful to young minds. Yet as McGuinness makes clear, discipline is pointless unless used in a technically effective manner, and we must learn how to teach the sounds of the language as an alphabet code that will give children the tools necessary to read, write and spell.

The National Reading Panel, for its part, has done the public no favor by bestowing its blessing indiscriminately on reading programs that are fundamentally at odds with one another and by refusing to recognize that, when reading philosophies are radically, structurally opposed, one will necessarily assume priority over the other in the classroom. Coles and his colleagues in the education establishment may believe that the debate is still between “flexible, democratic, and creative” whole language and “rigid, authoritarian, and mindless” phonics, but for the rest of us, the facts of the world and human nature teach that a measure of intellectual discipline—sometimes rigid, sometimes authoritarian, sometimes even mindless—is no bad thing, and that in teaching the young to read, it turns out, is a very necessary thing.