The Best of Both Worlds
Jack sits at his desk and ponders the worksheet exercises his teacher has handed out: Circle all the words on the page that contain the same "at" sound. After 20 minutes, Jack's teacher collects the worksheets and hands out a new set with a similar task. When that work is completed, the class moves on to a science lesson.
In Jill's 1st-grade classroom, the children sit on the floor around their teacher as she reads aloud to them. Upon finishing the story, Jill's teacher instructs the students to return to their desks and spend 20 minutes writing about the topic of their choice.
The two teachers use two markedly different approaches to teach their young charges how to read. The first, explicit decoding instruction, is known in shorthand as phonics or basic-skills instruction. The second is called whole-language instruction. Together, they have incited the great reading war that has raged in education circles for more than two decades.
But mounting empirical research strongly suggests that many children need explicit phonics instruction, leading experts to believe that a combination of the two styles may be the most effective way to teach the beginning reader. As convincing as the studies are, though, many educators still disavow the findings. They claim that traditional research methodology is ill-suited to gauging children's ability to read.
"We do know how to teach kids how to read in the beginning grades," says Steven A. Stahl, a professor of reading education at the University of Georgia and a researcher at the National Reading Research Center. "But at the same time, we're lost in this endless debate. It's kind of like Sisyphus who pushes the rock up to the top of the hill, and all of a sudden, it falls down again."
Balanced instruction combines the best elements from phonics instruction and the whole-language approach. That is, children are explicitly taught the relationship between letters and sounds in a systematic fashion, but they are being read to and reading interesting stories and writing at the same time.
The critical juncture comes during the early elementary years, specifically kindergarten and 1st grade. And students who fall behind early have a hard time catching up. Down the line, of course, intensive-and usually costly--remediation can help some of them. But why not teach them to read well earlier, when success is more likely, rather than later?
Whole-language people thought--indeed still think--they were doing just that, particularly reaching the students with the greatest risk of read-ing failure. They embraced the theory that kids learn to read the way they learn to talk--naturally. To foster this natural ability, whole-language educators surrounded children with real literature and nonfiction and threw out the old basals and worksheets that frequently manipulated words, offered nonsensical passages, and turned kids off to reading.
Early studies also indicated that children recognized words better in context rather than in isolation. So whole-language practitioners emphasized comprehension and meaning in text--what the whole word, the whole sentence, the whole paragraph, and the whole book meant--and guided children to look for semantic and syntactical clues to decipher unfamiliar words. And they had children write, edit, and write some more to buttress the connection between reading and writing and forming meaning.
That doesn't mean they ignore phonics altogether. Whole-language theorists also believe in providing students with mini-lessons. For instance, a whole language teacher might help a student sound out words, but the instruction is given in context to individual students on an as-needed basis.
Several comparative studies during the 1980s did seem to favor the whole-language approach. Since then, though, the research--at least the studies that fall into the conventional realm--tends to point to a more balanced approach.
Much of this research centers on what is called phonemic awareness--the understanding that sounds make up language. Numerous studies indicate that children who acquire phonemic awareness, which is different from phonics, will become more skilled readers.
"If you don't have that understanding and look only at words as containing meaning, you will have an awfully hard time," says Jean Osborn, a longtime researcher at the Center for the Study of Reading at the University of Illinois at Urbana-Champaign.
The experts assert that phonics without phonemic awareness is inadequate. In phonics instruction, a teacher may tell a child, "the letter 'f' makes the sound you hear at the beginning of the word 'fish,"' Connie Juel, a professor of education at the University of Virginia, writes in The Leadership Letters, an occasional series from the Silver Burdett Ginn publishing house. "To a child without phonemic awareness, there is no 'beginning' sound in 'fish."'
Phonics and phonemic awareness are not ends in themselves, though some whole-language purists have claimed as much. "Research shows that phonics training and training in phonological awareness facilitates later reading comprehension, not just letter and word recognition," says Keith E. Stanovich, an applied-psychology professor at the University of Toronto.
In New Zealand, where children start school a year earlier than they do in the United States, researchers at the University of Auckland conducted studies in which 5-year-olds in whole-language classrooms received instruction in either explicit phonemic awareness or process writing, a trait of whole language. Both groups made significant gains in recognizing letter-sound relations, the investigators found, but the recipients of phonemic-awareness training also improved their spelling.
In a second experiment, 17 children were taught phonemic awareness and letter-sound connections, such as alliteration, rhyming, and blending sounds. A second group of 17 was trained to categorize semantically. For example, while playing flashcard games, the children were supposed to distinguish among words that spelled out animals or musical instruments. The third group of 17 received no special instruction. On all measures of phonemic awareness, reading, and spelling, the group that received the phonemic-awareness training scored higher than the other two groups. In fact, the group that received semantic training and the group that received no special training showed no difference in scores.
For the past decade, a half-dozen university research centers have been working with the National Institute of Child Health and Human Development to identify the causes and seek remedies for learning disabilities--most of them reading impairments that affect some 10 million, or one in five, children.
In testimony last year, G. Reid Lyon, the director of the institute's research project, told a U.S. Senate committee that most reading disabilities stem from a deficit in the most basic level of the language system--the phoneme.
The researchers expect to gain more insight through new technology that allows them to view brain activity while children read. Preliminary results indicate that the areas of the brain where phonological processing occurs behave differently in children with reading disabilities.
"About 75 percent to 80 percent of the population probably will respond to most any type of teaching method," Lyon says. But the others need explicit instruction. "Quite a few of the youngsters probably became disabled by poor teaching," he adds.
Karen R. Harris' daughter came close to being one of those children with a reading disability. At the end of kindergarten, Leah's teacher suggested that she might have a learning disability, which surprised Harris and her husband, both special-education professors at the University of Maryland.
They took their little girl to a specialist who ran her through a battery of tests. Leah performed at or above grade level in all indicators except her ability to decode words. Then, Harris recalls, the specialist asked if her daughter was enrolled in a whole-language program. "I think all that is wrong with this kid is nobody has taught her the code," the specialist told the parents.
Mother and father spent six weeks during the summer providing their daughter with intensive phonics instruction. Now in the 4th grade, Leah reads at the 6th-grade level.
Most language experts now doubt that children learn to read naturally. Barbara R. Foorman, a professor of educational psychology at the University of Houston, explained her reasoning in the October 1995 School Psychology Review. "Humans are biologically specialized to produce language and have done so for nearly 1 million years. Such is not the case with reading and writing. If it were, there would not be illiterate children in the world. Yet, nearly all children in all societies develop a language."
Emerging research has also cast doubt on the whole-language concept that children learn words as they encounter them in context. Stanovich, for one, found that it is the poor readers who are more likely to depend on context to cue them in.
Other studies have shown that predicting words in context is successful only about 10 percent of the time.
In the jargon, a lot of the talk is about "automaticity." Experts suggest that if early readers devote all their time struggling to recognize words rather than using the cueing strategies they learned through phonics instruction, they won't have the mental energy to comprehend what they are reading.
Eye-movement studies have shown that good readers process nearly every word and all letters in a word--findings that contradict the whole-language devotees' belief that readers skip over letters and words. Skipping letters and words can lead to mistakes, researchers say. For example, a reader who bypasses letters and uses syntactical cues would end up with a different meaning if he encounters 'house' in a passage but sticks an 'r' in instead of the 'u.'
"The only times you use context is at the end to check it; does it make sense," says Bill Honig, the superintendent of education in California when the state switched over to a literature-based--and some would say, whole-language--framework.
"I started, if anything, from the other side [of the debate]," says Honig, who has written a book about teaching from a balanced perspective. "I changed my mind about a lot of things." For instance? "I could have sworn I skimmed."
Many researchers also question just how helpful whole-language instruction is for at-risk students. In an analysis of comparative research through 1989, Stahl and his associates could not find a single study that showed whole-language instruction produced significantly better results among children from lower socioeconomic backgrounds.
If anything, the researchers say, disadvantaged children coming from literature-poor homes are more likely to need the explicit instruction than their middle-class peers.
But middle-class children may need direct instruction, too. Stahl points to research from the 1950s when Israel used what resembled whole-language instruction. The students appeared to be doing well until a large influx of poor Arab students enrolled, and the failure rates zoomed to 50 percent. That's when educators realized that the middle-class parents had been taking it upon themselves or had hired others to teach their children how to read--a step the poor parents had not taken.
"I suspect that is what is going on now," Stahl says. "Parents are taking it on on their own."
Karin L. Dahl of Ohio State University and Penny A. Freppon of the University of Cincinnati conducted one of the few recent comparative studies that does give whole language the edge. The pair looked at kindergartners and 1st graders in four basic-skills and four whole-language classrooms in two cities. Most of the children came from low-income families.
Both groups gained awareness of and experimented with letter-sound relations. But the whole-language learners showed more insight about what they were reading and tended to be more independent readers; after completing one book, for instance, they would pick up another.
Students in the whole-language classrooms did write more than children in the basic-skills classrooms, but they showed no appreciable difference in the kinds of writing they produced.
And while whole-language learners read and wrote for their own purposes, the students in the basic-skills classes thought of reading and writing as schoolbound activities.
Within the whole-language classrooms that Freppon studied, though, the teachers were also providing rich skills instruction, says Michael Pressley, a professor of educational psychology at the State University of New York at Albany. They just weren't providing it upfront. "We're still trying to figure out which of the balanced approaches is appropriate," he says.
Pressley and two of his colleagues, Ruth Wharton-McDonald and Jennifer Mistretta, tracked students and teachers in nine 1st-grade classrooms in four school districts in the greater Albany area from December 1994 to June 1995. They selected the classrooms based on supervisors' recommendations of outstanding and typical teachers.
The researchers looked for three specific achievement indicators: how actively engaged were students in academic activities; how difficult were the books they were reading at the end of the year; and how well they expressed themselves, developed a topic, and included traditional writing conventions in their writing.
In the classrooms where students were highly engaged in their academic pursuits and were reading and writing well, the researchers found the teachers followed an "exceptionally balanced" teaching strategy. In both the whole-language classrooms where skills instruction was downplayed and in the basic-skills classrooms, however, the students often appeared bored, and their reading and writing achievement fell far short of the students in the balanced classrooms.
But ultimately, what could prove the most damaging to the whole-language movement is mounting evidence that children immersed in whole-language instruction are no more motivated to read than children who have learned to read via more traditional methods.
Researchers at Georgia Southern University set out to size up student reading attitudes. In the first experiment, they tested 485 1st through 5th graders in whole-language classes and 433 students in the same five grades who received sequential structured skills-based instruction. The results showed no meaningful difference between the two groups.
The data from the same basic-skills district were used in the second experiment, but the researchers selected a school that adhered more strictly to the tenets of whole language to make up their other group of 713 students. The second experiment bore out the results of the first.
"These studies provide no evidence that the wholelanguage philosophy offers inevitable advantages over traditional instruction in building students' attitudes toward reading," conclude Michael C. McKenna, Beverly D. Stratton, Martha C. Grindler, and Stephen J. Jenkins in the March 1995 Journal of Reading Behavior.
All the evidence, however, is not making believers of whole-language proponents. They contend that conventional research sets up artificial experiments. "The research is skewed by its design," says Kenneth Goodman, a professor of language, reading, and culture at the University of Arizona. "They follow a paradigm of research that says you have to conduct experiments and have to control all the variables but one. You produce this anomaly that some kids seem to be more phonemically aware than others."
So in the past few years, the wholelanguage researchers have concentrated on descriptive research--case studies and ethnographies--or studies that have come out of other cultures or relate to second-language learning.
"The kind of research going on now is instructional," says Jerome C. Harste, a language-education professor at Indiana University. "How can we set up those natural learning environments in classrooms and what happens when we do."
Their counterparts working in the more conventional realm say that while this type of research is useful, it isn't a sufficient underpinning on which to base beginning reading instruction.
"Whole language is a symptom of a field that does not have at its core the use of scientific logic and public safeguards that you have in other professions," says Douglas Carnine, a professor of education at the University of Oregon. "I can't imagine any other profession promulgating a practice that ends up harming literally hundreds of thousands of children."
But Harste, Goodman, and other whole-language supporters don't put much stock in the conventional research.
Says Harste: "Research is not innocent. It's not the place to go to find truth. You've got to look at who did the research and what are the ideological beliefs of the person doing the research."
Whole-language boosters even go so far as to suggest that researchers have become the unwitting pawns of the conservative and religious right whom, they believe, want to eradicate all features of whole language.
They also cite another reason for dismissing the research that touts balanced instruction: Teachers see for themselves what works in their classrooms and what doesn't. "One of the things that I think is deeply resented by the researchers is the advent of large groups of knowledgeable teachers who aren't coming to them hat in hand and saying, 'Teach me,"' Goodman says. Instead, he argues, the teachers are saying, "'Listen to me, sonny; I have a few things to teach you."'
But this kind of argument is precisely what supporters of a balanced approach maintain has confused the issue. As much a philosophy as it is a pedagogy, whole language champions teacher-controlled instruction and child-centered learning, as well as other noble tenets. Whole-language practitioners fear they may lose this autonomy if they are required to add phonics instruction to their repertoire.
"They are strictly independent from issues of the nature of the knowledge and processes involved in reading and learning to read," write Marilyn J. Adams, the author of the oft-cited Beginning to Read: Thinking and Learning about Print, and Maggie Bruck, an associate professor of psychology and pediatrics at McGill University in Montreal, in the Summer 1995 American Educator. "Only by disentangling these two sets of issues can we give either the attention and commitment that each so urgently deserves."