Most of the ways economists talk, if they were translated into English, would sound plausible enough to noneconomical folk like farmers and poets and business executives. The talk is hard to follow at first, in the usual way of specialized talk, because the culture of the conversation makes the words arcane. But the people in an unfamiliar conversation are not Martians. Underneath it all (the economist's favorite phrase), conversational habits are similar. Even mathematical models and statistical tests, which sound alien to literary ears, grow out of ordinary talk. Under scrutiny they reduce to words that even an earthling might use.

All the conversational devices of economics, whether words or numbers, may be viewed as figures of speech. They are all metaphors, analogies, ironies, appeals to authority. Figures of speech are not mere frills. They think for us. Someone who thinks of a market as an “invisible hand” and the organization of work as a “production function” and coefficients as being “significant,” as an economist does, is giving the language a great deal of responsibility. It seems a good idea to look hard at this language.

If the economic conversation were found to depend heavily on its verbal forms, this would not mean that economics is “not a science” or “just a matter of opinion” or some sort of confidence game. Good poets, though not scientists, are serious thinkers about symbols; good historians, though not scientists, are serious thinkers about data; good scientists, too, use language.