Sound Over Sight in Reading

A new way of teaching reading, already in wide use in the U.S., is challenging the "look-say" method that took over the field beginning 40 years ago. Look-say, best known through the "Dick and Jane" readers, counts on sight identification of whole words, using pictures as clues, and brings in phonetics only gradually. The new method, without being a throwback to McGuffey, is centered on phonetics, freely uses picture clues and--most significantly--puts to work on a broad scale the theory of programmed learning.

The switch is the work of California Linguistics Expert Maurice William Sullivan, 40, whose interest in language goes back to his hitch as a marine teaching German to U.S. Navy officers during World War II. Many a degree later (B.A. and M.A. in English at Yale, Ph.D. in linguistics at Madrid, B.A. in Spanish at Puerto Rico, M.A. in Spanish at Middlebury), he took up reading theory at Hollins College and Stanford and then retired to a hilltop in California's Santa Cruz Mountains to develop his books. Distributed by McGraw-Hill, they are now used by 200,000 children in some 2,000 schools in all 50 states.

Consistent Sounds. Sullivan's system requires children to spend their first eight weeks learning the alphabet from their teacher. But they are not taught all the sounds of all the letters. His "structural linguistics" approach keeps children from the confusing phonetic inconsistencies of the language (the 40 different sounds conveyed by the letter a, for example) until they grasp the fact that in general, letters correspond to sounds.

Thus Sullivan students at first learn only the sounds associated with the consonants f, m, n, p, t, th, and the short forms of the vowels a and i. Then they learn, purely by sight, a few such basic words as yes, no, on, the. With this equipment, when they turn to their readers they can read short sentences, sounding out such words as ant, man, pin, thin. In the first seven books, which average first-graders will complete in a school year, they learn roughly 375 words by sounding them out, often using clues offered by simple cartoon-like drawings. None of the words involve a phonetic conflict, such as the long o sound in doe, dough, row or sew.

Sullivan's beginning vocabulary is drawn from the 5,000 to 15,000 words that most five-year-olds already speak and understand, even if they cannot read them. Sullivan contends that most reading primers are compiled from word lists that have no logical basis; each list came from a survey of the most used words in older readers, and all went back to McGuffey, "who must have obtained his list from God." Sullivan and a research team financed by Encyclopaedia Britannica Films, Inc. compiled their lists instead by exploring the world of the five-year-old. "A little kid is very sane," says Sullivan. "He just won't pay any attention to something not intrinsically interesting."

Knowledge in Bits. All of this is organized according to the learning theories of Harvard Behavioral Psychologist Burrhus Frederic Skinner (TIME, March 24, 1961). Skinner taught pigeons to play pingpong by breaking the action into tiny steps, immediately rewarding each correct step with a grain of corn. This led to the idea of giving children knowledge in atomized "bits," and testing each bit immediately by an easy leading question. When the student responds with the right answer, he gets a glow of pleasure--his grain of corn. The technique requires some mechanical device (often a teaching machine) to hide the printed answer until the student is ready to compare it with his own. Sullivan's solution is to print answers on the left side of each page, which children can cover with a cardboard slider. So as not to reveal answers to upcoming questions, the left-hand pages are printed upside down, and the child flips the book over after reaching the back page, works through the book again.

Holding the child's interest is vital to programmed instruction, since each child works alone at his own pace. Breaking reading down to simple steps that lead a child progressively toward more difficult words, yet do not bore him, was Sullivan's greatest problem. His first attempts failed badly. A member of his team at the time, Psychologist Allen Calvin, tried programming a Superman story, found that it held kids' interest about nine times longer than the reading program. Even a programmed version of a Sears, Roebuck catalogue did five times better. "We were terribly discouraged," recalls Sullivan.

Nevertheless, Sullivan kept quizzing kids, found that they could be coaxed to chuckle first over impish-looking drawings of red ants, a fat man, even a thin pin, later over a frolicsome poodle named Nip and a red-headed moppet named Walter--all illustrating stories with plots that children found engaging. Laboriously trying out frame after frame on children and rejecting those that led either to boredom or too many wrong answers, Sullivan's team completed 21 textbooks--three series of seven, roughly intended for the first three grades.

On to Labyrinth. The first book requires a child to look at a drawing, then answer such questions as "Am I an ant?" or to circle the right word in distinguishing between mat and man. New letters are introduced, and by Book 7 he can handle such words as sandwich, haystack and yesterday. Each step requires the child to either make a yes-no choice, select the words to complete a sentence, or fill in a blank in a sentence. By Book 21 he has been introduced to all the toughest exceptions to the phonetic rules of English. His reading vocabulary totals 2,892 words --including deceitful, labyrinth and ridiculous--and he is reading action-packed stories about Greek mythological heroes. Teachers normally limit the children's reading sessions to no more than a half-hour. In that time a child responds affirmatively about 100 times. And each response, says Sullivan, means that "learning takes place inside the learner."

In careful comparisons with children taught by conventional methods, programmed reading has consistently come out on top. The average child who has completed one year of the Sullivan program, for example, scores in the third-grade level on the standard Gates Reading tests--even though the Gates tests are based on the different vocabulary of conventional primers. The brightest 10% in most Sullivan first-grade classes read at fifth-grade level.

A major reason for the speedup is that the step-by-step procedure helps a teacher spot precisely what is puzzling a child--and the method frees her "to give such children help without holding up others. It also gives her a chance to cope with one of the most worrisome facts facing every elementary teacher: the broad range in mental age (at least four years in a typical first-grade class) among her students. But the real key to the program's success, in Sullivan's view, is that in his books "the kids were the authors."

This file is automatically generated by a robot program, so reader's discretion is required.