Back in the early 1900s, when the professors of education were working overtime to find “scientific” justification for changing reading instruction in American schools from alphabetic phonics to the look-say, sight, or whole-word method, many studies were done to see what type of effect the new teaching method would have on children’s reading ability.
One study done by Myrtle Sholty, published in the February 1912 issue of the Elementary School Teacher, revealed that the two methods of teaching reading produced two different types of readers: objective and subjective. The alphabetic-phonics method produced fluent, accurate, objective readers while the sight method produced impaired subjective readers who guessed at words, omitted words, inserted words, substituted words, and mutilated words. The sight readers’ lack of phonetic knowledge put them at a distinct disadvantage. They were unable to accurately decode the words since they looked at them as whole configurations, like Chinese characters, with no connections to the sounds of the language.
Reading researcher Geraldine Rodgers, in an unpublished manuscript on the history of reading instruction, states that Sholty’s experiment merely confirmed what had been discovered in 1903 by German psychologist Oskar Messmer, who had identified the two types of readers. Rodgers wrote:
When William Scott Gray [future editor of “Dick and Jane”] published his summary of American reading research in 1925, which has been the foundation for all “histories” of “reading research” ever since, he “naturally” omitted Messmer’s German work, and “accidentally” misreported Sholty’s research in his brief summary so that it was no longer recognizable concerning either its nature or its conclusions.
Sholty was reporting on her tests with three little girls half-way through second grade, so the tests must have been done before 1912, probably after February, 1911. Of the three second-grade girls, two were reading words in parts, for sound, but one was reading only whole words for meaning. However, all three little girls at the University of Chicago experimental school were “helped” by context guessing, which was obviously necessary because of the small amount of phonic training used at the experimental school. Sholty specifically referred to Messmer’s research and noted that her research results were in line with his conclusions.
In 1914, psychologist Walter F. Dearborn, who reviewed the Sholty study, wrote about Messmer’s observations:
The chief differences between these types [of readers] are said to be that the objective readers have a rather narrow span of attention in reading, but see accurately what they do see, and seldom guess or “read into” the material perceived, and that the subjective readers have a wider span, are influenced more by words lying in indirect vision, depend on relatively meager visual cues such as large word wholes, and that they are more likely to misread because of the large apperceptive element which they supply to the reading.
And so it was well known by the top psychologists involved in creating the new look-say or sight reading programs that these whole-word instruction methods produced inaccurate subjective readers. Despite this, the professors proceeded to devise and publish the textbooks based on this very defective methodology.
Another very significant study, published in the November 1914 issue of the Elementary School Teacher, was done by Clara Schmitt, an assistant in the department of child study at the Chicago board of Education. She analyzed the errors made in oral reading by two groups of children: one mentally "defective," the other "normal." She wrote:
Defective children are sometimes capable of acquiring very large visual vocabularies, but show themselves quite deficient in perceiving phonetic relationships. Children of the first grade may be expected to acquire the simplest phonetic elements of the English language. The child who can obtain a visual vocabulary with facility, who gains a perception of the simple phonetic values, and who learns to combine them correctly for the independent learning of new words is considered a favorable reactor so far as the subject of reading in the first grade of the public schools is concerned.
The "normal" children chosen for the test were average-to-good readers, aged 7 to 11. The "defective" children were between the ages of 10 and 16 who had been in special rooms for "defective" children for at least one year. Since at that time the official policy of the Chicago public schools was to teach children to read phonetically, both the "normal" and "defective" children had been taught the same way. While the "normal" children learned to read phonetically with ease, the "defective" children had problems.
But what is particularly interesting in this study is the discovery that the "defective" children made very different types of errors, even though they had all been taught to read phonetically. Miss Schmitt writes:
The errors in pronunciation made by the normal children in this and the second reading test were always in favor of a word which had considerable visual or phonetic resemblance to the correct word. The errors made by the defective children with the first selection which was perfectly familiar to them in content, at least, were absurd as far as visual or phonetic values were concerned, but were calculated to fill in the context. The defective child reads, for instance that the fox saw a vine with berries [instead of grapes] on it. Because of the great prevalence of this type of variation the performance of the defective group cannot be compared with that of the normal.
In other words, it was easier for the "defective" child to substitute a word that fitted the context than decode the word accurately, which means that the "defective" children were reading like nonphonetic sight readers. And that is the way "normal" children are being taught to read today! For example, in Evaluation: Whole Language, Whole Child, a book explaining the wonderful world of whole language, the authors write:
The way you interpret what the child does will reflect what you understand reading to be. For instance, if she reads the word feather for father, a phonetic-oriented teacher might be pleased because she’s come close to sounding the word out. However, if you believe reading is a meaning-seeking process, you may be concerned that she’s overly dependent on phonics at the expense of meaning. You’d be happier with a miscue such as daddy, even though it doesn’t look or sound anything like the word in the text. At least the meaning would be intact.
In other words, a whole-language teacher would prefer that a child entirely misread words and guess at what the word should be based upon the context of a sentence! But even the early advocates of the whole-word method realized that they would have to teach some phonics. This was obvious from an analysis made by Josephine Bowden in 1912 of how children learned a “sight vocabulary.” She found “no evidence in any of the cases studied that the child works out a system by which he learns to recognize the words. That he does not work out phonics for himself comes out quite clearly in the transposition test. Furthermore, only once did a child divide a word even into its syllables.” Her conclusion:
Under the methods of instruction employed with this class as outlined above, it appears that these beginners in reading have after two months or more of instruction secured a sufficient concept of the general appearance of a very limited number of words to recognize them as wholes, that in doing this they made use of only very general cues or points of differentiation between words and have not noticed the finer points of distinction between words and parts of words. It appeared very doubtful to the experimenter whether, under this method of teaching words as visual wholes, the pupils would of themselves, have come to make this latter necessary analysis with much success. Without some foregoing analysis and subsequent synthesis, the differences between words are not great enough to be recognized merely from the total visual appearance. The early instruction of phonics may supply, in some measure, this analysis.
Contrast what Josephine Bowden wrote in 1912 about the necessity of teaching phonics in a look-say reading program with what whole-language guru Frank Smith wrote in Reading Without Nonsense in 1985:
Children do not need a mastery of phonics in order to identify words that they have not met in print before.... Once a child discovers what a word is in a meaningful context, learning to recognize it on another occasion is as simple as learning to recognize a face on a second occasion, and does not need phonics. Discovering what a word is in the first place is usually most efficiently accomplished by asking someone, listening to someone else read the word, or using context to provide a substantial clue.
The difference between Josephine Bowden and Frank Smith is that Bowden came to her conclusions after observing real children in a real classroom, whereas Smith was writing from theory alone. What is important about the three early experiments conducted in Chicago is that they taught us three important facts about reading instruction.
The Sholty experiment confirmed that the two teaching methods — phonics and whole-word — produce two different types of readers. Phonics produces accurate objective readers; whole-word methodology produces error-prone subjective readers. The Schmitt experiment revealed that today’s "normal" children who are taught to read by look-say make the same type of errors that "defective" children, those incapable of learning to read with phonics, make. In other words we are training children to make reading errors! And the Bowden experiment proved that without some phonics, the whole-word method was dismally inadequate as a reading instruction program.
But we’ve also learned much more about the reading process in the 80 years that the whole-word method has been the dominant form of reading instruction in American schools. A good many of the children who were taught to read with Dick and Jane still struggle to read as adults. We now know that the look-say method can cause dyslexia, which can only be cured in special education by expensive remediation. But that’s the way the progressive educators want it. And that is why with all the talk about education reform, there is no real effort to change the way reading is taught in American schools.