Friedman Quote

Sabine H found this great quote by Milton Friedman:

“Truly important and significant hypotheses will be found to have “assumptions” that are wildly inaccurate descriptive representations of reality, and, in general, the more significant the theory, the more unrealistic the assumptions (in this sense). The reason is simple. A hypothesis is important if it “explains” much by little, that is, if it abstracts the common and critical elements from the mass of complex and detailed circumstances surrounding the phenomena to be explained and permits valid predictions on the basis of them alone. To be important, therefore, a hypothesis must be descriptively false in its assumptions; it takes account of, and accounts for, none of the many other attendant circumstances, since its very success shows them to be irrelevant for the phenomena to be explained.”

Lubos has now also weighed in.

It is no exaggeration to say that the quote is paradoxical sounding, and deliberately so. It comes from Friedman's 1953 essay The Methodology of Positive Economics (ppg 13-14).

Bee and Lumo each take a crack at explanation, and I think Bee is more or less on target. Lubos is, well, Lubos - this time with Bayes Theorem. Read it if you're interested.

Friedman himself is more helpful. His next two paragraphs read:

The converse of the proposition does not of course hold: assumptions that are unrealistic (in this sense) do not guarantee a significant theory [no shit Sherlock!]
...
To put this point less paradoxically, the relevant question to ask about the “assumptions” of a theory is not whether they are descriptively “realistic,” for they never are, but whether they are sufficiently good approximations for the purpose in hand. And this question can be answered only by seeing whether the theory works, which means whether it yields sufficiently accurate predictions. The two supposedly independent tests thus reduce to one test.

The less paradoxical version is not only more coherent, it is also quite different. In particular, it doesn't pretend that the unrealism of the hypotheses is important to the theory but merely states that the theory abstracted must be predictive.

Consider a comparative case in physics: frictionless classical motion as taught in elementary freshman physics. The theory is indeed useful when the approximation is good enough "for the purpose at hand" but it's not that the assumption of no friction is false that makes the theory important, it is the fact that there are cases when the assumption is very nearly true. The "unrealism" => "importance" business is just a bit of squid ink thrown up to cloud the next issue he wishes to discuss, the assumption of "perfect competition" in classical economics.

The question in that case, as in all such cases, is whether the "assumptions are sufficiently good approximations for the purpose at hand."

Impressively unrealistic assumptions are no virtue, but insufficient fidelity for the purpose at hand is indeed a vice.

UPDATE: There is a sense of the word "unrealistic" which permits one to make sense of Friedman's quote. We sometimes consider "realistic" and "idealized" to be opposites, and if you equate "unrealistic" with "idealized" then the comment takes on some rationality. More idealized models have fewer moving parts, and hence are easier to compute with. This is an unusual and somewhat unnatural meaning for "unrealistic," which more commonly is used to mean:

Not compatible with reality or fact; unreasonably idealistic

From the Free Online Dictionary

The above meanings turn Friedman's F-twist into nonsense and the logical fallacy I mentioned previously. I thought Friedman was a native speaker of English. It seems odd that he would make such an obvious semantic error.