Columbus, OH, July 26-29, 1998
By Dr. Patrick Groff, Professor of Education Emeritus, San Diego State University
According to its Program notes, this conference was designed to provide its attendees information about the characteristics of "exemplary phonics instruction." To that effect, sessions were held on the following topics:
To make a critical examination of the details of the information presented at the OSU Summer Phonics Institute ’09, it first is necessary to identify the source of the information offered by its speakers, as proof of the validity of their contentions. For instance, at this meeting professor of education Richard Allington referred to "research" that he maintains supports his conclusion that "decodable" words are a relatively unneeded feature of beginning reading teaching. A decodable word in a written text is one for which a beginning reader previously has been instructed to recognize, by learning to sound-out each of its letters, and to blend these speech sounds together into a proper serial order. This blending results in an approximate pronunciation of the word. After this, it is found that novice readers can readily infer the correct pronunciation of the word.
However, it no longer is feasible to refer to the term, reading "research," with the assumption that it holds a common meaning for all who come into contact with it. Quite to the contrary, there are at present two remarkably different research paradigms used in discovering the characteristics of exemplary phonics programs: experimental research and qualitative research.
Until fairly recently, the most popular of these two research models was the experimental (empirical or scientific) approach. For example, in a typical implementation of the scientific method of reading research, it is hypothesized that inclusion of distinctive phonics teaching variables into a reading instruction program will produce significantly greater word recognition learning by students in this program, than otherwise is possible.
A matched-group of students, and teachers with the same general level of instructional ability as the experimental teachers, serve as a control group in this empirical investigation. The control reading program is the same as the experimental one except it does not include the unique phonics teaching variables found in the latter.
The progress of students in learning to read of both the experimental and the control groups in the study then is measured with a standardized reading test. The objective data gathered in this regard is subjected to sophisticated statistical analysis to determine which reading program was the more effective, i.e., to make sure that differences in reading scores between the two groups did not happen due to chance factors.
It is clear that this piece of experimental research is replicable. That is to say, it is designed so that researchers in the future can repeat its procedures to determine if its findings can be duplicated. After a scientific study has been replicated a satisfactory number of times, and produces similar findings in each instance, these date are judged to be valid. At that point, recommendations for their application to reading instruction are offered. Modern statistical science also has developed a mathematical process in which data from experimental studies on a given topic, that have dissimilar designs, may be acceptably combined and analyzed. This recent advance in statistics provides even more reliability for experimental research findings than traditionally was the case.
Qualitative Research is the second major form of educational inquiry. Qualitative studies on phonics information are conducted with remarkably different goals in mind than those set for experimental research. For example, qualitative research does not pose a hypothesis about the best way to teach students phonics information. It does not design and conduct a closely controlled examination of this issue, nor carry out an inquiry of that presumption in a neutral, disinterested, and skeptical way. It does not collect nor statistically analyze numerical data, nor make claims it generates serviceable generalizations about the best way for all students to learn phonics information.
To the contrary, qualitative research on phonics instruction begins and proceeds with the investigator interacting with students in an informal, nonintrusive, and ingratiating fashion. The objective here is to observe students in natural classroom settings, as they are learning phonics information and how to apply it to decode written words. The qualitative researcher will write nonnumerical accounts of what he/she has witnessed (rather than to run tests on students). These accounts are official records of what he/she believes took place as students acquire phonics knowledge. The researcher’s observations of the learning processes that occurred in this regard are considered to be more critical than are products of students’ performance that may be gathered.
To that effect, narratives, anecdotes, memoirs, case histories, results of interviews and questionnaires, audiovisual records, and the like, take the place of test scores. The qualitative researcher must be endlessly creative and inventive, even intuitive in gathering these kinds of evidence, and especially when determining their relative importance or pertinancy. Nonetheless, it is attested that classroom teachers make effective qualitative researchers.
Proponents of qualitative research insist that the kinds of evidence they gather about students’ phonics skills are superior to those gained through experimental queries. It is said the latter do not have the capacity to satisfactorily detect students’ learning processes. Particularly held up to scorn in this respect are standardized reading tests, the kind customarily used in experimental research. Their scores are misleading, it is held, since they do not take into consideration the total natural context in which students’ reading behavior occurs. Moreover, standardized tests, given only seldomly, supposedly cannot be authentic measures of reading ability because students’ reading skills are in a constant state of flux.
It also is maintained that standardized tests wrongly presume there is an external reality about students’ knowledge of phonics information waiting to be discovered, that is independent of students’ personal and eccentric perspectives of it. Qualitative research views the validity, reliability, and objectivity of evidence about phonics skills as being highly problematic and personal in nature. That is to say, reality about phonics information varies unpredictably according to how individual students personally perceive its existence. Thus, the only reality about phonics information putatively is that constructed by students involved in the process of dealing with it, qualitative inquiry insists.
With these theories in mind, qualitative research rejects the experimental research assumption that it is necessary for research on how students best acquire phonics skills to be as value-free and unbiased on the researcher’s part as is possible. In qualitative research, broad allowances are made for such partisanship. It is held that as long as qualitative researchers are aware that their peculiar dispositions, prejudices, or temperament are influential factors in their study of phonics teaching, the merit of their investigation of it will not be compromised. Therefore, the qualitative study of students’ acquisition of phonics information makes no pretense at dealing with students’ reading behavior in a dispassionate, or otherwise disinterested fashion. The empathy and attachment shown by qualitative researchers toward their student subjects in a study of phonics acquisition does not result in a distorted accounted of this learning, it is argued. To the contrary, it is proposed that it has a salutary effect on discovering the truth on this matter.
Almost all the findings on how students best learn phonics information that is produced by members of the Whole Language movement are gained through qualitative research. The wholesale adoption by Whole Language advocates of qualitative research, and their rejection of the experimental variety, is understandable. None of the unique principles nor practices of Whole Language is corroborated by relevant experimental evidence. Findings from the latter source consistently contradict those from qualitative studies, and vice versa.
Whole Language advocates thus avoid a potentially embarrassing loss of prestige of their scheme by conducting qualitative research about the workings of their unorthodox instructional innovation. The avoidance of experimental research for this reason clearly is imperative. The former research is a more convenient choice for Whole Language for yet another reason.
Whole Language purports to be a "philosophy" of teaching reading and writing. It thus is largely ideological in nature. Therefore, the qualitative research that Whole Language promoters conduct is aimed at explaining and confirming its underlying philosophy. This research is not designed to critically examine the Whole Language philosophy for its potential faults. It is assumed there are none. The only valid research finding thus is one that compliments the novel Whole Language doctrines and practices, it is said. In this manner, Whole Language maintains itself as an irreproachable and unassailable instructional approach. Proof of the success of this impregnable defensive stand is the fact that none of the multitudinous complimentary accounts of Whole Language that are published cite a qualitative study that found fault with its original theories and operating procedures.
The distinctive nature of Whole Language instruction, as opposed to teaching based on experimental evidence, has never been better summarized than by one of the leaders of this movement, professor of education Carole Edelsky (Educational Researcher, November 1990). Whole Language holds "a different view of education, language, and learning; uses different discourse; maintains different values; and emanates from a different educational community." It is based on "a major philosophical framework [deconstructionism] and a particular political ideology" (socialism). It offers different answers to questions such as "What is reality? Where do facts come from? What is truth? How should power be distributed?" (away from those presently in authority to the poor and people of color).
One of the main attractions of Whole Language to teachers is its claim to be the most humane form of reading instruction available. It is true, in this regard, that Whole Language is more concerned with upgrading students’ dignity, freedom, and power to decide how and what they will learn, than with how fast they learn to read competently. It also captures teachers’ fancy with its promise that once it is instituted into schools, teachers will become the exclusive judges of how well students have learned to read. Teachers also have been taken in by Whole Language’s false assertion that it alone favors having students read much high quality literature, and frequently engage in written composition and in editing and sharing it with others.
Many teachers who subscribe to Whole Language also doubtless are impressed with the purported simplicity of the scheme. The guiding principle of Whole Language is that students best learn to read in the same natural, informal manner in which they previously learned to speak at home, as preschoolers. Thus, in bona fide Whole Language classes, direct, systematic, intensive, comprehensive, and early teaching of a prearranged sequence of phonics skills is greatly de-emphasized. Instead, students are "immersed" in written language (e.g., are read to aloud as they "follow along" in identical texts). From this experience students readily are able to infer how much phonics information they personally need to learn in order to develop effective written word recognition ability, it is held. Group teaching of prearranged phonics information also is rejected on the grounds that each student has one of the large number of unalterable learning styles, most of which are incompatible with explicit phonics instruction.
To be come genuine Whole Language advocates it also is necessary for teachers to accept several other empirically unverified assumptions about phonics instructions. For example, that:
A survey of the academic activities and publications of the reading instruction specialists who were chosen to address the attendees of the OSU Phonics Institute reveals that they all are members of the Whole Language movement, in good standing. These are professors of education Richard Allington, Diane DeFord, Mary Jo Fresch, Amy McClure, John McCracken, Gay Su Pinnell, Patricia Scharer, and Jerry Zutell. No notable reading instruction specialist, who honors experimental research on phonics instruction, and thus who is a negative critic of Whole Language, was invited to speak at the Institute.
Accordingly, the Institute was misnamed. In line with its purpose, it should have been called the OSU Summer Phonics (According to Whole Language/Qualitative Research) Institute," or "The Institute on Exemplary Phonics Instruction As Interpreted by Whole Language Proponents and Qualitative Research." Any attendee at the Institute who anticipated hearing both sides of the current international debate over the best way to teach phonics information, and the relative importance of instruction of phonics knowledge in beginning reading programs, doubtless was sorely disappointed by the Institute’s program.
Without doubt, the proceedings of the Institute were unsatisfactory and misleading for anyone who was unfamiliar with the particulars of the present widespread debate about phonics teaching. These would be beginning or relatively inexperienced teachers, new school board members, parents, the general public, state legislators who have responsibility for overseeing reading instruction, and members of the business and commercial community.
Compounding this problem is the fact that papers presented at the Institute offer a combination of some empirically irrefutable facts about phonics teaching, mixed in with information about it gleaned from qualitative studies that is irreconcilable with experimental evidence. One would have to be fully informed about the details of the present controversy over phonics teaching to be able to sort out these two kinds of information into their appropriate categories. (One exception to this fault was the copy of G. Reid Lyon’s "Why Reading Is Not a Natural Process," Educational Leadership, March 1998, provided at the Institute.)
It is not the purpose of this critique to determine the exact extent to which this comingling of contradictory experimental and qualitative evidence took place. Nonetheless, generally speaking only seldom in the publications presented to Institute attendees can one find a page whose content can be judged to be based entirely on qualitative as versus experimental findings, or vice versa.
Not only is it proper that these publications be negatively criticized as not making clear whether their sources were qualitative as versus experimental research. They also must be faulted as raising frivolous and spurious objections to the experimental evidence that discredits Whole Language teaching. A prime example, to that effect, was the copy presented at the Institute of Richard Allington and Haley Woodside-Jiron (1998), "Thirty Years of Research in Reading: When is a Research Summary Not a Research Summary?" This material is a chapter in the book, In Defense of Good Teaching, edited by Kenneth S. Goodman, a co-founder of the Whole Language movement.
Here Allington and his colleague attack the recent review of experimental research on students’ learning to read conducted by the National Institute of Child Health and Development. This animosity toward the survey is not surprising. It concluded that empirical evidence does not support the Whole Language approach to reading instruction.
Allington and his partner make three supposedly trenchant complaints about the NICHD survey: (1) it improperly included studies of students who had experienced difficulties in learning to read, and/or were "at risk of [experiencing] such difficulties"; (2) "the NICHD studies were not successful in improving students’ [reading] performance"; and (3) the investigations inspected by the NICHD did not reflect one way or the other on "the importance of decodable text." In actuality, these objections respectively are (1) irrelevant, (2) inaccurate, and (3) obtuse.
Considering the review of experimental research on reading made by Rudolf Flesch in 1955, and the ones conducted by Jeanne Chall in 1967, 1983, and 1989, it is clear that the survey of the empirical evidence on reading carried out by the NICHD was an appropriate one. There also are more than fifty other recent critical analyses of the scientific evidence of reading that bear witness to this fact.
Although they offer no more than imitation proofs "that phonics [teaching] is being oversold," to schools, Allington and his associate are not deterred from contending that a dastardly "disinformation campaign" against Whole Language of a "political" nature, is afoot in the nation, managed by the NICHD. They chide the Whole Language movement for not carrying out an "effective political public relations campaign" to decry the "legislative or regulatory restrictions on the nature of beginning reading materials" that the NICHD survey stimulated. To this effect, they bemoan "there have been no position papers, no targeted mailings, no telephone campaigns, no media blitz, no glossy flyers, no news conferences, no news releases, no lobbying, no nothing" assembled and carried out to defend Whole Language teaching. Underlying this protest is the proposition that qualitative research evidence gathered by Whole Language enthusiasts should not be required to defend itself in the marketplace of educational ideas. The manner in which the OSU Summer Phonics Institute presents only one side of the current debate over phonics teaching echoes that sentiment.
"Perhaps the lesson to be learned is that learning to read is a political topic," Allington and his cohort conclude despairingly. In so saying, they inadvertently reveal a truth that may signal the demise of Whole Language. In California, for example, Whole Language became more popular than in any other state. As a consequence, students in this state devolved into the least capable readers in the nation. When this reading instruction disaster finally reached the attention of the California legislature, it held comprehensive hearings to determine its cause. Both sides of the great debate over phonics teaching were given ample opportunities to offer their reasons.
The side that honored instruction based on experimental research argued the California reading teaching catastrophe was the result of Whole Language teaching. On the other hand, Whole Language proponents claimed it was due to insufficient funding of the schools, white racism, a lack of school libraries, the public’s disrespect for teachers, attempts to privatize the schools, a lack of diversity in school enrollments, too much standardized testing, the effects of drugs and gangs on students, insufficient teacher pay, attempts to make English the official language, and so on – anything but the fact that Whole Language is an empirically discredited form of instruction.
In this dispute, California lawmakers came out on the side of the negative critics of Whole Language. It no longer is lawful in this state to conduct the worst excesses of Whole Language teaching in public schools. California lawmakers thus have made sincere efforts to afford students the best chance possible to learn to read, knocking down decades of ideological barriers to that goal put in place by Whole Language teaching.
Summary and Recommendations
The Ohio legislature (or that in any other state) and school districts cannot legitimately decide how phonics instruction should be conducted in its public schools without the kind of dull hearings on this matter held in California. The OSU Summer Phonics Institute ’98 therefore is not sufficient for that purpose. As has been described in this discussion, the Institute presented only a singly, highly partisan position regarding phonics teaching. This is the one promoted by the experimentally discredited Whole Language movement, and by the qualitative research findings that the movement produces.
Such information alone is not adequate for deciding how phonics should be taught, i.e., what are the characteristics of exemplary phonics instruction. Accordingly, it is imperative that a second phonics institute in Ohio be organized and conducted called, "The Institute on Scientific Phonics Instruction." Attendees at this meeting would be presented conclusions about phonics instruction drawn from the empirical evidence. If a certain practice in phonics teaching was supported by experimental findings, but not by those from qualitative research, teachers would be advised to adopt it. To the reverse, a teaching practice recommended by qualitative research would not be acceptable if contradicted by empirical evidence.
In short, to make rational decisions as to how phonics information is best taught, it often is necessary to make forced-choices between teaching recommended by experimental as versus qualitative evidence. The assumption that some reading instruction specialists make in this regard, that incompatible evidence from these two competing sources somehow can be balanced, melded, or otherwise combined is worse than merely unsound. It poses a severe handicap to the development and establishment of exemplary phonics instruction programs. Without such superior phonics programs, students will be denied what schools are basically designed to deliver: full opportunity for each student to learn to read.
Dr. Patrick Groff, Professor of Education Emeritus, San Diego State University, has published over 325 books, monographs, and journal articles and is a nationally known expert in the field of reading instruction.
Source: Reading Research Foundation