Weighing Claims of 'Phonics First' Advocates
Does research support a "phonics first" approach to teaching reading, as this method's proponents often argue?
This question is not entirely answerable by factual evidence or statistical data; the issue is partly a matter of opinion--a question of what we value. For those who value the ability to construct meaning from texts over the ability to identify words in isolation, the phonics-first approach does not appear to be the best strategy. Indeed, emphasizing direct, systematic phonics instruction in the primary grades may do more long-range harm than good.
Despite claims to the contrary, all educators agree that children need to learn that English is to a significant degree an alphabetic language and that they need to develop a functional grasp of basic letter-and-sound patterns. What is at issue is whether phonics should be taught and tested in isolation, as a prelude to reading interesting texts--the phonics-first view--or whether phonics strategies should be developed more gradually, in the context of reading and writing materials that interest students.
The debate has been fueled recently by federal legislative actions and by the distribution of controversial new documents about reading. An amendment to the literacy bill passed by the Senate in February would allow federal money to be used for training teachers in phonics instruction. This legislation was influenced by a Senate Republican Policy Committee document--entitled "Illiteracy: An Incurable Disease or Education Malpractice?"--suggesting that the cure for illiteracy is "the restoration of the instructional practice of intensive, systematic phonics in every primary school in America!" This paper, in turn, cites the summary of Marilyn J. Adams's study, Beginning To Read: Thinking and Learning About Print, newly released by the Center for the Study of Reading and supported by the Education Department's office of educational research and improvement.
From the point of view of fact, we should ask: Are such recommendations supported by evidence? Is the evidence fairly represented? Have phonics-first advocates considered all the relevant research or only a select sampling? And perhaps more important, from the viewpoint of values: What are researchers measuring when they claim that systematic phonics produces better reading "achievement"? Might there not be more important things to measure, if our long-range goal is to help students become literate enough not merely to read signs and labels but to participate fully in American life?
The absurdities and falsehoods of the Republican policy committee's alleged "study" might seem laughable if this document had not been distributed to governors and state legislators as well as members of the business community and educators; or if Senator William L. Armstrong of Colorado, the committee's chairman, had not recently had it read into the Congressional Record; or if, using the document as a rationale, Mr. Armstrong had not persuaded his colleagues to pass the aforementioned amendment.
The simplistic argument that systematic phonics will cure illiteracy is not, however, the only reason for educators' concern about this document's wide distribution. Most policymakers lack sufficient background knowledge to question the adequacy or veracity of the evidence offered, much less the pronouncements made without any evidence.
In fact, misinformation abounds in the paper. It claims, for example, that there are basically only two ways to teach reading: phonics and "look and say," less derogatorily known as the sight-word method. Other labels used to describe the latter approach, the document continues, include "whole language." Nothing could be further from the truth.
In a sight-word approach, children read stilted primerese--"stories'' with new vocabulary words repeated at least five times in a selection--and they may be drilled with flash cards containing frequently used, "basic" words. In whole-language classrooms, children read and reread favorite rhymes, songs, and patterned stories with repeated phrases, sentences, and stanzas--not single words repeated in unnatural contexts. Gradually, students learn to recognize many words in isolation as well as in the familiar contexts, and gradually--but with teacher assistance--they develop the phonics knowledge they need to read.
A sight-word approach uses a part-to-whole strategy, and in this respect, it resembles phonics. By contrast, whole-language instruction moves from wholes toward parts, often in a single day's activities.
Ignoring this crucial distinction, the "Illiteracy" paper lumps together look-and-say, whole-language, and language-experience (also different from a sight-word approach), and then claims that this method "is currently used by nearly 85 percent of the schools in the United States." Even many advocates of systematic phonics would be surprised by this assertion. For example, Jeanne Chall--cited approvingly in the document--points out in the 1983 update of her 1967 book Learning To Read: The Great Debate that "soon after the late 1960's, basal reading programs introduced more phonics, and test developers began to test decoding skills in standardized achievement tests for the early grades."
To determine the validity of the claim that the "look and say" method is widely used, educators, parents, policymakers, and other interested citizens need only examine the teachers' manuals and pupil workbooks of the basal programs that have been the mainstay of the reading curriculum in the vast majority of our schools. At least from the early 1970's to the late 1980's, virtually all of these programs have introduced phonics in the primary grades, typically in kindergarten. This may not be the kind of "intensive, systematic phonics" advocated by the drafters of the document, but it is most certainly phonics. It is also more phonics than most children need--and far too much for many children.
We must ask, then, whether the sources cited in this paper are any more reliable than the document itself. In fact, many of these studies are deficient. And the conclusions that popular summarizers draw are subject to a criticism implicit in Richard Turner's research review in the December issue of Phi Delta Kappan: that any "advantage" for early systematic phonics continues for only a short while and is evident only when reading "achievement" is measured by standardized tests that typically place a higher premium on word-attack and word-identification skills than on comprehension of connected text.
Though the summary of Ms. Adams's report is much more scholarly, her study actually contributes no new evidence to support the effectiveness of systematic phonics. She does cite two studies of the effects of Follow Through, a direct-instruction program that places heavy emphasis on phonics, but the conclusion that heavy phonics instruction leads to improved reading is suspect.
For instance, close examination of a study reported in the American Educational Research Journal reveals that, while 5th and 6th graders who were in the Follow Through program during their primary years scored much higher than comparison groups on a test of decoding skills, they scored considerably lower in reading comprehension. Ms. Adams reports this result simply as an overall difference in favor of the Follow Through group. Similarly, scrutiny of a study of high-school seniors does not clearly indicate that either phonics or direct instruction can be credited with the various kinds of superiority Follow Through students demonstrated in comparison with their peers.
Not only does the research typically cited in favor of phonics fail to support this approach, but the contrary evidence is rarely even acknowledged by phonics-first advocates.
Some of this research is admittedly too new to have reached mainstream journals. Much of the more recent work is described by Diane Stephens in a report forthcoming from the same Center for the Study of Reading that published Ms. Adams's study. For example, in what the researcher claims was the first quantitative comparison of a whole-language with a code-emphasis approach, Helene Ribowsky compared two classes of kindergartners and found that the whole-language group showed greater gains and scored better on all post-tests, even on the Metropolitan Achievement Test's letter-recognition and phoneme/grapheme test for consonants.
In a study involving at-risk 1st and 2nd graders, Carole Stice of Tennessee State University and Nancy Bertrand of Middle Tennessee State University found the whole-language students to achieve better results according to all measures, including the reading portion of the Stanford Achievement Test.
These and other recent studies suggest that, in comparison with children in whole-language classrooms, phonics-first students may not always do better on tests of isolated skills admin4istered in the primary grades.
But determining which approach can produce the highest scores on standardized tests in the primary grades should not be the central issue--much less the only issue.
Results from the National Assessment of Educational Progress reading tests suggest that we are succeeding rather well at what has all too often been our major aim: improving scores on tests of "basic" skills, including literal comprehension and low-level inference. But we are not adequately stimulating the development of students' ability to read or reason critically.
For example, in the 1984 assessment, only 39 percent of the 17-year-olds tested could find, summarize, and explain relatively complicated information, and only 5 percent could synthesize and learn from specialized reading materials. The comparable figures for the 1988 assessment are 41.8 percent and 4.8 percent respectively.
Given such results, it is not surprising that major education groups are encouraging us to consider what we mean by literacy, what kinds of literacy the next generation will need to develop, and how instruction, assessment, and research can promote the kinds of literacy8deemed important. The concept of literacy implicit in the report of the English Coalition conference, for instance, is much broader than that reflected in our measures of "achievement."
If our long-range goal is to help students develop sufficient literacy to participate fully in our society, heavy emphasis on phonics and other low-level skills may subvert our intentions. Students who have trouble with skills-work are often assigned more of the same, given fewer opportunities to read, and in general viewed as incompetent at reading. Such treatment lowers self-esteem as well as desire to read, thus contributing to the likelihood that these students will develop only marginal literacy at best.
Research, then, must look primarily not at test scores but at factors such as these: Are students reading to comprehend, rather than just say or identify words? Are they developing a flexible repertoire of reading strategies, including techniques for synthesizing and evaluating what they read--and write? Do they view themselves as good readers and writers who can enjoy and profit from various kinds of materials? Do they voluntarily read and write outside of school? Are they developing the ability to think critically and creatively through language?
In short, are they developing the attitudes and habits of independent, self-motivated, lifelong readers and writers?
Whole-language classrooms foster these habits and values to a far greater degree than more traditional classrooms--especially those emphasizing phonics first.
Vol. 09, Issue 27, Page 32