Fundamentally, an investor can obtain profit from a stock in two ways: by receiving dividend payouts from the company or by earning capital gains through selling the stock after it has risen in price. In today's market, the latter reason has come to dominate investor motivation. In order for a stock to rise in price, other people must buy it. It is therefore "rational" for an investor to try to predict the inclinations of the rest of the existing and potential investor community. The question of whether a company is, in some objective sense, "good" or strong or profitable is really irrelevant to the investor, except as it affects others investors' attitudes toward its stock. If other investors want it, then it's valuable.

True, a share of stock is by definition a percent ownership in a company. But this is a legal abstraction rarely relevant in the real world. (Bankruptcy is the principal exception.) It is possible for a stock price to move independently or even inversely to the performance of the company. There is no reason why a drop in quarterly earnings must lead to a decline in stock price, except that each stockholder fears that others, reacting to the information, will sell their stock and wishes to escape the resulting depreciation of his assets. If investors could be persuaded that others would not sell their stock, the price would not fall. The issue, essentially, is the actions of other investors, not the status of the company.

Despite this fact, most stock analysts devote less time to an attitudinal investigation of the investing community than to an analysis of the so-called fundamentals: price/earnings ratio, dividend levels, etc. It's not hard to see why. "Fundamental" research yields a more concrete, quantitative result and lends the financial professions greater legitimacy. Furthermore, in times of relative market stability, research into the fundamentals will lead to the same conclusions. All the other analysts are looking at the same data and the investing public is relying on their advice.

But methods of valuing stock are only as effective as the number of investors who subscribe to them. Those who discarded their price-equity calculations were scolded by purists for not sticking by the tried and true markers of rationality. But why should they, when the only reason that the P/E ratio was helpful in predicting stock prices was that everyone else was using it as well? Similarly, if the P/E ratio has become less helpful, it's not because of new technologies or changing economic dynamics, but simply because enough people were persuaded that the old valuation methods were no longer where it's at.

How is this related to my former professor's provocations? It suggests that the way we have been discussing the stock market and investor behavior is inaccurate. Markets are "rational" in the sense that investors' behavior is directed toward their self-interest and they will generally make reasonable choices to promote that interest. The problem occurs when the word "rational" is used as a substitute for "efficient," or to suggest that stock price should conform to other measurements of the value of a company. Often stock prices will reflect these external value measurements, but only when investors are collectively possessed by the belief that it should. The discarding of P/E ratio and other traditional valuation methods may not be a good thing. It may contribute to the market's volatility, exposing the investor to greater risk and companies to greater instability. But the Internet stock boom did not make the market more or less "rational."

If anything, last year's boom -- and this month's roller coaster -- have only exposed the extent to which market movement is, and has always been, based on perception, psychology and convention.

Visit Salon's newest site Salon Audio for free spoken word recordings from your favorite authors