The other day I opened a letter from an unknown address in Kitchener that was addressed to me. Dear Physics Expert, it began; and as I read this salutation, my eyes misted over and my heart sank faster than a lead balloon.

Physicists are not, it is true, besieged by mail, but unsolicited letters that begin Dear Physics Expert have, in my experience, an alarming tendency to deal with conspiracy theories of suppressed publications and invariably end with wild-eyed claims about perpetual motion machines humming away in nearby basements.

I glanced through the letter quickly and noticed that the second page concluded with pencilled illustrations of geometric figures in a childlike hand -- a still more troubling sign.

As I prepared to place the letter at the top of my burgeoning "to be more rigorously examined once every conceivable pressing issue of the day has been dealt with" (otherwise denoted as the Mephistophelean snowblower file), I noticed that the signature of the author had a peculiar, uncertain quality about it, as if the very act of signing the name was somehow an evolving practice.

I turned back to the opening line of the letter. That's when I discovered, to my amusement, that my correspondent was an eight-year-old Grade 4 student named Tim who was writing to me about a theory he had created.

Needless to say, I read on.

Tim called his theory "perspectivity." Concerned about how to categorize pictures of shapes placed inside each other, he developed a classification procedure to assess whether or not, in his judgment, certain relative proportions between the two figures held. Those figures where these proportions held, he called "inper;" otherwise he called them "outper."

Truth be told, I had a hard time precisely following Tim's line of reasoning, but what puzzled me the most was his use of the word "theory."

Tim had developed some way of classifying certain shapes that he created, which was clearly both creative and precocious.

However, a vital aspect of the usual scientific thought process was missing -- namely, the fundamental "why" questions: Why these particular shapes? Why should one care about these particular proportional relationships?

In other words, Tim had given a lot of thought to how to classify some types of objects in certain constructed circumstances, but far less to what could be done with these classifications or what these classifications would tell us about the physical or mathematical world.

That is fine and reasonable, you might say -- after all, Tim is only eight years old. Quite so. But the point is not to berate Tim for failing to develop a coherent theory of the universe or unveil deep mathematical truths; the issue is how one might best educate clever, inquisitive students like Tim to become better aware of the fundamental principles of rational inquiry and deduction.

After all, the concept of a scientific hypothesis and objective theoretical framework is one of the most significant accomplishments of modern times, but it is not self-evident: it must be somehow learned.

How to do this, exactly? It is not clear. At the family dinner table that night, I decided to conduct a small experiment. I would ask my eight-year-old son and seven-year-old daughter what sound was until they began, in the best Socratic tradition, to build a theory.

My son, ever keen to impress me with his knowledge of basic science, informed me quite breezily that sound came in waves. Ah yes, I replied, but waves of what? Waves of wavy things, replied my daughter equally confidently. Indeed, I persisted, but what was waving where?

The two of them shrugged and changed the subject, clearly tiring of the game.

"What do you think is going on?" I pressed further, "give me a possible explanation."

But this held no interest for them whatsoever. They just wanted me to provide the answer.

And so, faced with preaching to them or changing the subject, I folded like a rotten banana and quickly took the easy way out. I talked about undulating air molecules and vibrating eardrums and how a speaker works. I did my utmost to fill their heads, for a moment or two, with facts. My experiment lasted 10 minutes or so and ended in abject failure.

From this I could only conclude one thing: teaching children how to think scientifically is hard. Perhaps I just need to have more patience. Or perhaps I need a better theory.