Learning To Read and Whole Language Ideology

by Jeffrey M. Jones, M.D..,Ph.D.

Probably no other subject provokes such bizarre interactions between teachers, school administrators, and parents as the selection of beginning reading programs in American schools. Many districts have "closet phonics teachers," who close the doors to their classrooms, get out their phonics-based teaching materials, and teach their children how to read using phonics. When a parent expresses chagrin over the lack of phonics in reading instruction or the discovery that his child cannot read, school administrators typically say, "There is lots of research to support our approach." The "approach" of course is an "eclectic reading program" or "whole language program." The educators never produce any concrete research to back up their assertions. But the parent is invited to produce research that proves the educators wrong. This is a rather stark role reversal. Shouldn't it be the education professional who is charged with producing research to convince the parent? Hopefully this essay will make life a little easier for parents caught in this predicament.

Reading is the process of constructing meaning from written texts(1). Phonics is concerned with teaching letter-sound relationships as they relate to learning to read (1,2). While English writing is based on an alphabetic code, there is not a one-to-one relationship between graphemes (printed symbols) and phonemes (speech sounds) they represent. "Cracking the code" denotes learning to associate printed letters with the speech sounds they represent. In discussing beginning to read, cracking the code refers to learning letter-sound relationships via the ability to apply phonics. When a child has learned to associate all specific printed letters with specific speech sounds, the code has been mastered, or cracked. The child then can arrive at an approximation for the pronunciation of most printed word symbols (1,2). "Phonics-first," "intensive phonics," "systematic phonics," "decoding," or "code emphasis" refer to reading programs that emphasize use of phonics at the inception of reading instruction and throughout the first 1-3 years of reading instruction. "Look-and-say," "whole-word," "sight-reading," "linguistic," or "psycholinguistic" refer to an approach to reading instruction taken by educators such as William S. Gray during the first third of the twentieth century. They wanted to turn schools away from "heartless drudgery." Gray advocated a look-and-say approach. He thought children would make more rapid progress in reading if they identified whole words at a glance, as adults seem to do. In its more conservative form, look-and-say involves teachers teaching a limited number of sight words to children before phonics analysis is introduced. In its severe form, educators backing sight reading seek to avoid phonics completely. I will use "look-and-say" to refer to reading teaching that minimizes structured use of phonics. One may hear the term "language experience" (2) used to describe a program for teaching reading. In its conservative form, such a program tends to work on speaking, listening, and writing skills as well as reading. Actually almost any good reading instruction program will interface in some way with the teaching of these skills. In its most severe form, language experience becomes the same thing as "whole language." As you will see later, the term "whole language" means different things to different educators. This is a common problem that parents face when educators talk. You frequently don't know what they really mean. Whole language can mean simply having printed material readily available to children to use and reading to kids to get them excited about reading. This is a concept with which anyone would agree. Whole language can also denote a belief system, an ideology, which, if used to guide reading instruction, can be quite destructive.

Basal reading programs are a complete package of reading materials (1). They provide an entire reading curriculum (summarized in a "scope and sequence" chart), instructional strategies for teaching reading (through teachers' manuals), a graded anthology of readings for children (readers or primers), and practice exercises (work sheets and workbooks). Basal reading programs are organized by grade level with most programs beginning at kindergarten and continuing through eighth grade. An entire basal reading program would make a stack of books and papers four feet high. To develop one of these programs, a large publishing company may invest up to $15 million. Needless to say many more millions are earned yearly selling these programs so that there are vested commercial interests in use of any particular program (1). The programs are heavily marketed, with five well known programs having 70% of the American market. (1,3)

Although educators often say teachers use whatever reading materials and techniques work for them in beginning reading instruction, in fact commercial basal programs drive reading instruction strongly. Studies have shown that basal reading programs account for 75-90% of what goes on during reading periods in elementary school classrooms (4). A number of classroom studies indicate that, for the most part, teachers follow instructional strageties prescribed in a given program's teachers' manuals and that students use the program's Readers and workbook materials (5).

Parents need to be able to ask probing questions about basal programs selected by their schools and about reading instruction in their childrens' schools. Becoming a Nation of Readers (BNR) is an excellent resource for interested parents (1). This 145-page monograph was prepared by the Commission of Reading of the U.S. Department of Education. Although it is the product of the Commission, its staff, and 35 consultants, it is quite readable. It clearly made waves in the educational establishment. Because the report was likely to be quite influential, the National Council of Teachers of English (NCTE) had difficulty bringing themselves to distribute it (6) when it was released in 1985 because BNR recommended the use of explicit phonics instruction in grades 1 and 2. The NCTE eventually produced a brief monograph to act as a reply to BNR (6), a reply that I found to be anemic in comparison to the BNR report. Unfortunately, it appears to me that the NCTE and the International Reading Association (IRA) have had a leadership over the past several years that is dominated by whole language enthusiasts. In my discussion later regarding whole language, I will reproduce some statements by Kenneth Goodman, who is past president of the IRA. The concensus of the experts who wrote BNR was that skilled reading is constructive, fluent, strategic, motivated, and a lifelong pursuit. Phonics plays a role in enabling students to develop their decoding skill to the point where it is automatic and requires little conscious attention.

Parents need to know that there is a group of educators who espouse a pure whole language ideology for reading instruction that rejects use of phonics and use of basal reading materials. I found it difficult to find a concise, logical explication of whole language. These folks tend to prove points by speaking in analogies. Because pure whole language is a set of beliefs to be acted upon and defended, I think it is best termed an ideology. Whole language adherents believe "language is whole" and that reading will happen if you just immerse kids in language. More specifically they tend to believe: (a) written language is language; (b) when a child is surrounded by speaking and naturally occuring writings ("literature" by definition) the child will learn to read "incidently" as a natural "personal" (psycholinguistic) and "social" (sociolinguistic) experience; (c) use of phonics and prepared basal reading materials to teach "skills" (used in a pejorative sense) will inhibit natural learning; (d) there is a large orally transmitted (teacher to teacher, researcher to researcher/teacher) and written body of research supporting (a) and (b); and (e) teachers who choose to use whole language to teach reading must be supported or society is infringing on their rights. Items (b) and (c) of this list appeals to Piaget's philosophy and cognitive learning theories in a vague sort of way. Item (e) seems to mix reading instruction into the teacher empowerment movement. I assert that as bizarre as these ideas seem, when one strips away the flowery rhetoric of whole language enthusiasts, the above five items constitute their credo. I will offer you several exhibits to support my summation of their belief system and then discuss each exhibit separately:

Exhibit A: Definition of whole language according to Diane Stephens in Research on Whole Language: Support for a New Curriculum (10):
1. Learning in school ought to incorporate what is known about learning outside of school.
2. Teachers should base curricular decisions on what is known about language and learning, should possess and be driven by a vision of literacy, should use observation to inform teaching, and should reflect continuously.
3. Teachers as professionals are entitled to a political context that empowers them as informed decision makers.

Exhibit B: Excerpts from "Whole Language: What's New?" by Altwerger et al. (11):
[1] The key theoretical premise for whole language is that the world over, babies acquire language through actually using it, not through practicing its seperate parts until some later date when the parts are assembled and the totality is finally used. The major assumption is that the model of acquisition, through real use (not through practice exercises), is the best model for thinking about and helping with the learning of reading and writing.
[2] Language acquisition (both oral and written) is seen as natural - - not in the sense of innate or inevitable unfolding, but in the sense that when language (oral or written) is an integral part of functioning of a community and is used around and with neophytes, it is learned "incidentally"...
[3] Little use is made of materials written specifically to teach reading and writing. Instead, whole language relies on literature, on other print used for appropriate purposes (e.g. cake-mix directions used for really making a cake, rather than for finding short vowels), and on writing for varied purposes.

Exhibit C: Excerpts from "Twenty Questions about Teaching Language," by Goodman and Goodman (12):
[1] Early in our misuse research, we concluded that a story is easier to read than a page, a page easier to read than a paragraph, a paragraph easier than a sentence, a sentence easier than a word, and a word easier than a letter. Our research continues to support this conclusion and we believe it to be true....
[2] It is through errors...that we've learned that reading is a psycholinguistic guessing game... The Hawaiian child who reads "He was one big fat duck" for "He was a big fat duck" is letting us in on the ability to read one kind of dialect and translate into dialect in order to comprehend....
[3] ...we can teach children letter names and the sounds letters represent and we can teach them words in isolation from the context of language, but we know that these methods do not lead children to read.

At first glance the points in Stephens definition appear rather opaque. But when viewed through the lens of my characterization of whole language ideology you can actually see what she is trying to say. I recommend Engelman's book War Against the Schools' Academic Child Abuse (13) for a more complete explication of criticisms applicable to Exhibits [B] and [C]. In any event, stripped of its rhetoric, Exhibit [B] says:
1. Written language is language.
2. Babies acquire language through actually using it, not through practicing its seperate parts.
3. Oral language is learned "incidentally" (as a aspect of doing something else).
4. Therefore, written language is best learned "incidentally" (a la cake-mix routine)
This represents distorted logic. First, written language is not language. It is a representation of a language, hopefully of a language known to the learner. When a baby starts out babbling, what language is that? Also, the fact that a six year old knows a lot more language than the two year old suggests that many specifiable "parts" and aspects were learned in four years. It has been estimated that by the time a child is six years old he has a 4,000 to 24,000 word speaking and listening vocabulary upon which to draw in learning to read (14,15). We don't start from scratch when we teach reading. We assume a basic oral language understanding. Do parents really want a six year old, who has mastered language and who knows much about word meaning and syntax from speaking and listening to be learning a language when he learns how to read or to be learning to read "incidentally?" I doubt it. Parents expect the teacher to direct the child's learning, which is to be quite specific. It does not involve learning a language but a code expressing a familiar language. The teacher may have 28 students to teach the very specific things they don't know about "written language."

Now let's look at item [1] of Exhibit [C]. Can the child read a sentence without being able to read the component words? Really? If the words are harder than sentences, and if sentences are made from words, what phenomenon lets kids transcend the more difficult unit (words) to get the easier unit (sentences)? If something is harder than something else, probably someone would fail to learn it. Engelman (13) believes there ought to be something called the "Goodman Syndrome." These children would fluently read stories, but when asked to read a page, they would make many mistakes. They would stumble horribly over individual words and find letter identification impossible. Goodman's assertion also reminds me of performer Woody Allen's joke about speed reading. "I took a speed reading course," he said, "and it really worked! I read War and Peace in 20 minutes! It was about Russia." Item [2] of Exhibit [C] seems benign at first glance. However, whether a child is Hawaiian or not, substituting words is not a good thing for him to do. If teaching practices actively encourage this behavior, we get something the Merck Manual, a standard medical reference text refers to as one of the signs of "primary reading disability" or "word blindness," namely, "the tendency to substitute words for those one cannot read." Finally there is item [3] of Exhibit [C]. With a single sentence all the research supporting the use of phonics is dismissed. Here Goodman is confusing factors important in motivating children to read with the skills important to allowing them to do so.

What is the research concerning whole language as opposed to traditional programs that emphasize use of phonics? A comprehensive review of whole language effectiveness was conducted by Stahl and Miller in 1989 (16). They looked at 5 projects conducted as part of a U.S. Office of Education study of first grade reading programs and at 46 additional studies that appeared as dissertations, transcripts of lectures, or journals, which they felt had sufficient data to permit a metastatistical analysis. They concluded "we have no evidence showing that whole language programs produce effects that are stronger than existing basal programs, and potentially may produce lower effects. The alternative, that whole language programs are too new to evaluate, also suggests a lack of evidence of its effectiveness." There is also the review of the theoretical foundations of the phonics-first versus whole-language approaches to reading instruction produced by Professor Vellutino of State University of New York in 1991 (17).

He summarizes his findings as follows:

The implications of the research for teaching children to read should be apparent. The most basic dictate seems to be that instruction that promotes facility in word identification is vitally important to success in reading. Accordingly, instruction that facilitates both phoneme awareness and alphabetic coding is vitally important to success in reading. However, there is nothing in the research that precludes the use of whole-language-type activities in teaching reading, such as use of context for monitoring and predictive purposes, vocabulary enrichment to imbue printed words with meaning, discussion that would encourage reading for comprehension, integration of reading, writing, and spelling to concretize the relationships between and among these representational systems, and so forth. [emphasis added] (17).
Finally, there are two large monographs which have reviewed the research regarding reading instruction comprehensively. Both of these, one written by Jeanne Chall (18) and the other more recently by Marilyn Jager Adams (19) have concluded that phonics instruction is of prime importance for reading instruction, especially for the first one to two years of instruction.

The whole language crowd has had a simple retort to the assembled research that makes their positon unsupportable - - they ignore it. Diane Stephens, in the monograph from which Exhibit [A] was taken extolled the pure whole language approach. She listed 38 papers which she believes represent research that supports the unfettered use of whole language. Of these papers she reviewed, only 8 appeared in periodicals where some peer review was possible. The remainder were unpublished masters or doctoral dissertations, technical reports, book chapters, abstracts in yearbooks, or transcribed lectures. Nearly all the studies she cites are "descriptive" of classrooms implementation of whole language or "case studies" of individual students learning to read. She claims 10 studies were comparative; however, when I examine her synopses of these works, only eight truly compare what at least their authors label "whole language" with "skill based" instruction. Of the latter 8 papers, only three compared outcomes of instruction in any way. The studies were short term, looking at the outcome of instruction for the year they were employed. They tended to use nonconventional means for assessing reading ability. Overall, they found little or no advantage for whole language compared to traditional programs. The total number of students in the combined variable and control groups for the three studies was < 200 students! This is the quality of data the whole language crowd expects parents to accept as "research." Stephen says that the most positive thing about the whole language classrooms she described was that each teacher "behaved as if the desired were actual." Each teacher "believed - - and was observed behaving as if - - the students were competent, sensible, and well-intentioned." Do you remember when Harold Hill in the Music Man sold musical instruments to the parents of River City for their kids? He told the parents he was teaching the kids to play by the "think method." If you understand "the think method," then you understand whole language. In my opinion, educators should obtain informed written consent from parents of children upon whom they want to practice this unproven technique.

You may say that I am much too virulent in my attack on the use of a purely whole language approach in reading instruction. After all, educators would not force teachers to use exclusively whole language as their principal teaching method. In fact they certainly have done so in many instances across the republic. Parents have heard anecdotes from teachers describing their teaching phonics "on the sly." But consider the "Reading Learner Outcomes" specified in 1992 by the Oklahoma State Department of Education in their official Oklahoma State Competencies, Grade One, pages 15-22:

The student attend[s] to the meaning of what is read rather than focusing on figuring out words....Uses context, pictures, syntax, and structural analysis clues to predict meaning of unknown words. Develops a sight vocabulary of high frequency words...Predict[s] unknown words...Uses predictions in order to read pattern books (stories with a repetitive element)....Uses fix-it strategies (predicts, uses pictorial cues, asks a friend, skips the word, substitutes another meaningful word)....The student will interpret a story from illustrations.
What is missing from this picture? The word "phonics" is nowhere to be seen. Woe be to the Oklahoma teacher who adopts the outcome: "The student shall have phonemic awareness!" The teachers of Oklahoma better hide their phonics worksheets. This is blatant imposition of a whole language ideology on Oklahoma schools. Even by the twelfth grade, the Oklahoma outcomes still admonish the student not to "focus on figuring out words." However, he must nevertheless demonstrate "a positive attitude towards self as a reader."

The State of California, often touted as being at the vanguard of educational innovation, implemented whole language standards in 1988. It was sold as a package to provide a quality education for all its students, including its "diverse learners." California's DPI audited the schools to ensure compliance and by 1992, roughly 90% of the California fourth graders scored near the bottom of all states participating in a National Assessment of Education Progress (NAEP) assessment of reading (20). Stanford's Michael Kirst commented, "We almost beat Mississippi - - but not quite. For California to say that is just devastating (21)." Proponents of the whole language initiative then pointed to the growing number of diverse learners to explain the state's poor performance, even though a primary purpose of the standards in the first place was to help diverse learners. In considering diverse learners, it should be noted that white fourth graders in California also scored near the bottom when compared to white fourth graders across the U.S.. What was the answer of the educators promoting whole language? They said, "The kids just need a stronger dose of whole language to fix their reading problems." California continued to promote a whole language approach. In March,1995, the state's own test showed that a majority of its 4th, 8th, and 10th graders failed to reach even minimally acceptable performance levels in reading and writing (22). Then, in April of 1995, results from the most recently administered nationally administered NAEP reading test showed that California's 4th graders ranked last among the 39 states that administered the test (22). Only the island territory of Guam, where students also took the test, had worse scores. At the time this essay was written, a proposed law requiring the state to adopt spelling books and phonetic teaching materials is making its way through the California legislature. The California experience clearly shows that when an idea takes hold within the educational establishment, reference to rational methods for evaluation may be tossed out the window.
(Return to Topics Overview)


1. Anderson, R.C., E.H.Hiebert, J.A. Scott, and I.A.G. Wilkinson (preparers). 1985. Becoming a Nation of Readers: The Report of the Commission on Reading. Washington, D.C.. U.S. Department of Education.

2. Bond, G.L., and R. Dysta. 1967 The cooperative research program in first grade reading instruction. Reading Research Quarterly 2:5 - 142.

3. Shannon, P. 1983. The use of commercial reading materials in American elementary schools. Reading Research Quarterly 19:68-85.

4. Educational Products Information Exchange. 1977. Report on a National Study of the Nature and Quality of Instructional Materials Most Used by Teachers and Learners. Technical Report No. 76. New York: EPIE Institute.

5. Anderson, L. 1984. The enviroment of instruction: the function of seatwork in a commercially developed curriculum. In G.G. Duffy, L.R. Roerler, and J. Mason (Eds.), Comprehensive Instruction: Perspectives and Suggestions. New York:Longman, pp.93-103.

6. Davidson, J.L. (ed.) 1988. Counterpoint and Beyond: A Response to Becoming a Nation of Readers. 1111 Kenyon Road, Urbana, Illinois: National Council of Teachers of English.

7. From "Mix and make." 1982. In T. Clyman and R.L. Venezky. Ginn Reading Program (level 3, unit 2). Lexington, Massachussetts: Ginn and Co., pp.36-37.

8. From "At the seashore." 1982. In A. Hughes, S.A. Bernier, N. Thomas, C. Bereiter, V. Anderson, L. Gurren J.D. Lebo, and J.A. Overberg. The Headway Program (Level B1, lesson 17). LaSalle, Illinois: Open Court, p.67.

9. Basic Reading (1-1 Book). 1963. J.B. Lippincott, p.103.

10. Stephens, Dianne. 1991. Research on Whole Language: Support for a New Curriculum. Katonah, New York: Richard. C. Owen Publishers.

11. Artwergen, B., C. Edelsy, and B. Flores. 1987. Whole language: what's new? Reading Teacher 41:144-154.

12. Goodman, K., and Y. Goodman. 1981. Twenty questions about teaching language. Educational Leadership 38:437-442.

13. Engelman S. 1992. War against the Schools' Academic Child Abuse. Portland, Oregon: Halcyon House. 215pp.

14. Seashore, R.H., and J.C. Seegers. 1949. How large are children's vocabularies? Elementary English 26:181-194.

15. Lorge, I., and J. Chall. 1963. Estimating the size of vocabularies of children and adults: analysis of methodological issues. Journal of Experimental Education 32:147-157.

16. Stahl, S.A. and P.D. Miller. 1989. Whole language experience approaches for beginning reading: a quantitative research synthesis. Review of Educational Research 59:87-116.

17. Vellutino, F.R. 1991. Introduction to three studies on reading acquisition: convergent findings on theoretical foundations of code-oriented versus whole-language approaches to reading instruction. Journal of Educational Psychology 83:437-443.

18. Chall, J.S. 1983. Learning to Read: The Great Debate. Second Edition. New York: McGraw-Hill.

19. Adams, M.J. 1990. Beginning to Read: Thinking and Learning about Print. Cambridge, Massachussetts: the MIT Press.

20. Education Week. September 22, 1993.

21. Kist, MN. October 27, 1993. The Oregonian .

22. Witt, H. May 14, 1995. Bad grades for new age education: low scores may lead California back to old teaching methods. The Chicago Tribune.

P.R.E.S.S., P.O. Box 26913, Milwaukee, WI 53226
Phone: (414) 453-8116, FAX: (414) 453-9442, E-mail: presswis@execpc.com

Return to PRESS Home Page