WHY IS IT THAT the reading public in America consists of 15 or 20 million people out of 226 million? Why is it that the really committed readers number no more than 2 or 3 million -- probably less than the number of illegal immigrants in the entire country?

One answer, one big answer, maybe the big answer lies in the way Americans are taught to read. All of us who didn't go to schools of education (and many who did) know that that way is absurd and demeaning. Twenty years ago kids learned to read from Fun With Dick and Jane, an almost totally mindless book. Things are worse now. Dick and Jane have gone -- to be replaced by Janet and Mark. The girl's name comes first now, which may be a trifling gain for one sex. The mindlessness has increased. If Dick and Jane were an insult to any child's intelligence, Janet and Mark are an outrage. This is how Janet talks:

"Come, Mark, come. Come here, Mark. Come here. Come here, Mark, come and jump. Come and jump, jump."

And Mark? That clever child answers, "Here I come Janet. Here I come. Jump, jump, jump."

Sach dialogue is boring, meaningless, and totally unlike the way real 5- and 6-year-olds talk. We all know this, and yet we mostly kind of shrug and accept it for our children.

Partly that's because we are in awe of science and of authorities -- and you can be very sure that Janet and Mark has many credentials from authorities in phonics and cognitive skills. Partly it's because too many of us believe a pernicious lie about ghetto kids. What this lie says is that the vocabulary of pre-primers must limit itself to about 100 words (the first four books in the Janet and Mark series employ a total of 78) because otherwise culturally disadvantaged children would not be able to cope. So it's no good lamenting that in 1920 American first graders read books that contained four times as many words as the books they read now. That, the lie says, is part of the price of integrated and democratic schooling.

Bruno Bettelheim and Karen Zelan have written a book that could quite possibly change all this. If I didn't know the enormous power of the education establishment, I'd say that would change all this. Because they know a better way to teach reading.

The way it works now, the whole effort goes into decoding and phonics. That's why the books are such teash -- to leave the teachers free to concentrate on teaching the children such things as how to distinguish between the sounds of a and sh. And all mistakes the children make are assumed to be failures in decoding. Take a real case. A little girl is reading a "story" about a monkey and a seashell. With no trouble, she reads:

Did Mit see the shell?

I need the shell.

See this shell.

See this sand. I will sell it.

But then, coming to the line "I will sell the shell," she first stumbles, then blocks, and finally reads it, "I will sell the sheet." Wrong! She is instantly corrected for her failure in decoding.

But what Bettelheim and Zelan know is that there's a lot more going on here than phonics and decoding. They know it because they had discussed that "story" with the little girl and her whole class. They know, for example, that no one in the class felt it made any sense for the monkey to sell the shell. "They all preferred a pretty shell to any remuneration it could possibly bring." Bettelheim and Zelan also know that in an earlier reader, the little girl had encountered a rat who got a clean sheet dirty. What she did was to import meaning to the meaningless story by having the monkey sell something that she herself would have been willing to get rid of. The books exclude meaning; the children bring it back in.

The right way to teach reading, the authors say, is first to use books that tell real stories, in language that children really use. And second, not to concentrate on correcting "misreadings," but to treat them as meaningful substitutions -- which will nearly always enable the children to unblock. "An interim phase of making errors more freely [will] give way first to spontaneous correction and then to reading as printed."

One more example. This involves a sixth-grader named Lillian, who has a severe reading disability. She is reading a real book, though one very young for her: Maurice Sendak's Where the Wild Things Are. On page nine, having read "wild" correctly twice, she reads it as "mild" -- and then gets flustered and stops. The very distinguished educator who was working with her attributed this to her "inability to focus on the written word" -- a mechanical explanation. Bettelheim and Zelan point out that she has no trouble at all focussing on a television screen for hours at a time, or on the ball in a game of catch, She was focussing just fine. What really happened was that she had just come to the scary part of Sendak's book -- and she was rewriting it to make it less scary. Where the mild things are.

Because these are complicated psychological arguments (the authors spend 15 rather ponderous pages on Lillian alone), it is nearly impossible to make them convincing in a short review. Let me assume your conditional assent... and quickly go on to some of the objections that educationists are sure to raise:

* It's easy for Bettelheim and Zelan to prescribe theoretical cures, but only the teachers in the field really know what teaching reading is like. Fiddlesticks. This book is based on hundreds of hours spent in the classrooms of eight different American schools by eight researchers (including the authors).

* What proof is there that culturally disadvantaged children could handle real books? Lots. They do, for example, in Austria, Switzerland, Poland, Russia, and Japan. Bettelheim and Zelan have studied the primers in all these countries, and analyse several of them in this book. Russian first-graders get 10 times as many words as American first-graders.

More cogent still, it turns out that the real-life vocabulary of American children who are being subjected to Mark and Janet is around 4,000 words -- and the most culturally deprived in the most somber settings still know about 2,000 words, all but a few of which they have taught themselves. Small wonder that "jump, jump, jump" fails to grab them.

* They don't allow for dyslexia. Oh, yes they do. They just note that true dyslexia is a rather rare condition.

* But even if all this is true (it is!), very few gradeschool teachers are trained in psychoanalytic interpretation. You'd have to have a Bettelheim and Zelan in every classroom for their method to work.

No, you wouldn't. The teachers don't have to figure out why "sheet" for "shell," or "mild" for "wild." This is just a bit of bravura Bettelheim and Zelan use in making their case. All the teacher has to do is acknowledge the substitution. Interpretation isn't necessary -- though occasionally the child will spontaneously provide it.

Most American Schools use puerile books like Janet and Mark, and, inevitably, mostly fail to produce interested readers. It is a reasonable hope that some primary teachers and some principals will at least read Drs. Bettelheim and Zelan. I think they'd learn a lot. And if they did, so might our children.