The NY Times debuted a new column today that, the description says, "will examine problems teachers encounter in the classroom and how they deal with them." That's great. I've often said that I'd like to see general interest publications explain to readers the intricacies of teaching, the dilemmas, the choices, the strategies. Such information, in my view, counters the widely held notion that teaching is something anyone with the right disposition and motivation can do. The column will be written by Susan Engel, director of the teaching program at Williams College. Her first effort is well-written, clear, specific, direct, compelling.

But (and this is a big but) the first column, in my humble, non-expert, biased and unbelievably naive and ill-informed opinion, totally missed the point. The column is about a kid who can't read. And rather than the teacher actually teaching the kid to read, she has the kid read to her dog. So that's the solution. More dogs in classrooms!!!

The teaching problem the column addresses is what to do with a nice, sweet, amenable third grader who likes to draw but is a slow, reluctant reader who stumbles over new words. Naturally, then, he hates to read aloud. The set up to the column says that the child had memorized words in kindergarten. "Most children," the column continues, "catch on fairly easily after this, figuring out how to use the meaning of words and sentences as a guide to greater fluency." This is the heart of the whole language philosophy--reading is natural and fun and when kids are exposed to lots of words they infer the "code" of how letters are connected to sounds and even if they don't they can learn the meaning of words by guessing and predicting. Problem is, this is a terribly inefficient process. Why not help the kids along by letting them in on the secret of the phonetic properties of the letters and letting them know that, in most cases, the code will tell you what the word is?

I'm not talking about "drills." (I could hear the whole language crowd starting to rumble in the background.) I'm talking about conversations starting in preschool with kids in which the phonetic properties of letters are pointed out. "Oh, Richard, you say you've drawn a picture of a red racecar?! That's great. Did you notice that your name starts with the same sound as "red" and "racecar"? Instruction can be more explicit as kids get a little older. The point is that kids need to be taught the phonetic code, it's NOT something that most kids infer naturally, even if some do. You can teach that code in many different effective ways. And, it goes without saying that kids need to read a lot, they need to talk about what they read, they need to listen to others read, including the teacher, they need to be engaged in lots and lots of conversation starting very early so that they develop a rich vocabulary, they need to write and so much more.

But if the problem is fluency (the ease and speed with which kids read, including intonation), which it was in this case, the answer is not more time spent reading aloud or, as many advocate, reading silently. The answer is helping the kid learn the characteristics of the words he is stumbling over.

What does this mean for journalists? What it means is that when teachers make the same argument Engel makes in this column, push them a bit on these points. Don't just accept the idea that reading is natural. Speech is natural. Reading is not.