Phonics Debate Linked
to Nature of Educational Research

To go by my e-mail, reading instruction is one of the hottest topics in the field of education.

On March 7, I wrote that some -- perhaps most -- children require explicit, systematic instruction in how sounds of the English language are represented by letters (that is, phonics instruction) in order to learn to read well.

"Well, duh" is probably the most common reaction people have to that statement, but that just shows how little they know -- or how much they know, depending on your point of view.

The voluminous mail I have received on the topic illustrates how difficult it is to talk about reading instruction without getting embroiled in battles fought with a passion usually reserved for religious warfare. And yet, despite the fact that the effects of these battles are felt in classrooms around the country, most people don't even know the battles exist.

One of the more thoughtful letters I received lays out some of the issues:

If you knew, you are one of the cognoscenti. For the rest, let me see if I can fill you in on the background to this letter and the conflict it poses.

The National Reading Panel was convened by the U.S. Department of Education and the National Institute of Child Health and Human Development in 1998 in answer to a request by Congress that they determine what research says about reading instruction. Congress was reacting to the scary fact that it is quite clear that at least one-third of American children do not read well enough to become fully functioning members of a sophisticated, technological society.

The 14-member reading panel was chaired by Donald N. Langenberg, chancellor of the University System of Maryland. I asked him why he, an experimental physicist by training, was chosen. One of the reasons, he said, was, "I know what good research looks like."

That, by the way, was an ideological gauntlet he had just thrown down.

The hard sciences, such as physics, long ago established ways to sort through which research is worthwhile, which worthless and which intriguing but not proven. One is to require that experiments be designed so that extraneous factors do not corrupt the results. Another is to ensure that experts in the field review the research. Yet another is to publish the results in journals so fellow scientists can replicate the experiment and either verify the results or dispute them.

These and other long-tested methods have allowed scientific knowledge to be developed and built upon, piece by piece, by scientists who can rely on the previous research of other scientists. These methods are not foolproof, but they have proved useful and reliable.

About a century ago, the medical sciences adopted these standards for their research, with the result that we can now use research to guide medical decisions. That doesn't mean the research provides all the answers to our ailments, but its slow accrual of knowledge means we no longer have to put all our hope in Dr. Whatsit's Magic Liver Potion to cure liver cancer and "women's complaints" the way our great-grandparents did.

Most educational research does not adhere to these kinds of scientific protocols. Much of it is anecdotal or not described in such a way that it can be replicated, or doesn't sufficiently filter extraneous factors. As a result, almost any educational practice can be justified by some piece of "research" or another. Long-suffering teachers have learned to harden themselves against the words "research shows," knowing that what will follow is likely to be questionable.

In other words, because educational research has yet to establish its version of scientific standards, we're still in the stage of hoping that Dr. Whatsit's Magic Reading Program might work.

As head of the National Reading Panel, Langenberg, in consultation with the other members, established criteria for which research the panel would consider reliable enough to guide reading instruction. To be considered, research had to be published in a peer-reviewed journal, meaning that experts had evaluated it before publication. Research also had to include enough information so that it could be replicated, and it had to follow a number of other standards familiar to scientists. (To see the full report, including the research criteria, go to www.nichd.nih.gov/publications/nrp/report.htm).

Using those principles, the National Reading Panel whittled the number of studies it considered reliable from 100,000 possibilities to about 40. Needless to say, this has not exactly endeared the panel to those who produced the 99,960 other studies on reading, and some have vociferously protested the criteria.

This selectivity also means that the National Reading Panel limited itself to a very few conclusions, since only a few could be backed up by research that met its criteria. One of those conclusions was that "Phonics instruction is most effective when it begins in kindergarten or first grade."

I don't know what the 30-page report referred to in the letter above is. I have a 58-page summary that does not claim that phonics instruction alone can improve the reading of older children.

The National Reading Panel bemoaned the paucity of good research on a whole range of topics related to reading. But, it said, enough research exists to begin building a knowledge of what constitutes good reading instruction. And one part of that is systematic, explicit instruction in phonics, particularly for children in kindergarten and first grade and older children who are having difficulty reading.

Phonics instruction is by no means what all children need, the panel said. They need to have stories read to them by fluent readers so they can understand the wonder of a good story. They need lots of conversations using sophisticated vocabulary so they can build their storehouse of words and ideas. They need opportunities to write poems and stories that are fun and interesting for kids.

The only really controversial part of the panel's recommendations, as far as I can tell, is that it called for explicit, systematic phonics instruction, a conclusion disputed by those who say that phonics instruction is too boring for children and too quirky to rely on because English words don't always follow phonetic rules. But those dissenters haven't been able to produce research that meets the reading panel's scientific standards.

What's interesting about the way the National Reading Panel did its work is that if its standards stick and are applied to all of educational research, that would represent a genuine turning point in education -- the beginning of a common base of knowledge similar to what has been built in the medical sciences in the last hundred years.

"It's a little like folk medicine," Langenberg said, likening the current state of educational research to the pre-scientific era of medicine. "Over the ages, people learned a lot -- that if you eat that berry there, you die, and if you brew a tea from this bark, it alleviates pain. That's real. That's real knowledge," akin to teachers knowing that they can give more individual attention to kids in smaller classes. "But it's not biochemistry. We're ready to make that transition."

He would even like to see an educational equivalent to the Food and Drug Administration, in which publishers of curriculum materials would have to prove their effectiveness before taking them to market.

Making sure instructional practices have a scientific basis wouldn't eliminate the judgment of teachers, any more than medical research has eliminated the judgment of doctors. It would, however, allow all teachers to draw on a common storehouse of knowledge, something they have never been able to do in the past. And that could take some time to build.

"It's the work of a generation," Langenberg says.

Home | About Us | About Phonics | Resources
Research | Topics | Reading Reform | Links | Search

Unless otherwise noted, you may copy and distribute any information on this site as long as The National Right to Read Foundation at www.nrrf.org is given credit. The National Right to Read Foundation is a 501(c)(3) publicly supported organization.