Russ Roberts interviews my teach Alex Rosenberg on the nature of economic explanation and the role of math in economic explanation here.
Some notes and thoughts.
Rosenberg’s 1983 paper exploiting the “prediction and control” picture of science was written just as the “Good Till” date on that late 19th & early 20th century “pragmatic empiricist” philosopher’s account of “science” was coming up, with the re-discovery of non-linear dynamic phenomena, the developing literature on the significance of the sensitivity of initial conditions, and the mathematical advancements of Mandelbrot and Lorenz, all of the stuff made famous by James Gleick’s 1987 book Chaos: The Making of a New Science.
At that point the “pragmatic empiricist” philosopher’s dreams of a grid-like Laplacean world of linear predictions and simple statistical randomness exploded on the ground. The world simply doesn’t come like that.
Those who developed Darwinian biology, such as Charles Darwin, where already aware of this. Natural selection is an open-ended process, essentially sensitive to initial conditions, typified by functional categories multiply instantiated by an open-ended array of physical instantiations, bounded only by processes taking place across historical time (see my paper “Insuperable Limits to Reduction in Biology” written for Alex Rosenberg and Larry Wright in 1992.)
Now Hayek made the same point about economic categories in his papers on “Scientism and the Study of Society” — economic goods, and human categories more generally, were multiply instantiated functional categories which unfolded in open-ended processes across historical time. This was all part of Hayek’s “filling in” of the detailed gaps in Menger & Mises’ account of the non-Laplacean, non-predictive nature of economic categories and economic explanation.
Let’s acknowledge first that most everyone comes to the party pre-loaded with a picture of “science” and “knowledge”, pictures inherited from the philosophical tradition and from the first page or two of a freshmen Chemistry or Physics textbook — and most of these pictures are completely wrong.
My teacher Alex Rosenberg came to the economic science literature pre-loaded with a picture of knowledge, science and explanation take from the mid-20th century tradition of Nagel & Hempel — the “post-positivist” tradition combining aspects of Mill, Hume & Mach with the new logic of Frege, Russell & Carnap. What came out of that was the “covering law” model of explanation. By the end of 1970s that “normal science” program in the philosophy of science had by and large played itself out. Some of the last gasps of that program involved final efforts to extend and apply the framework to some of the “special sciences” outside of freshman physics, Michael Ruse in Darwinian biology and Alex Rosenberg in economic science, see e.g. Rosenberg’s Microeconomic Laws: A Philosophical Analysis, 1976.
By the time of Rosenberg’s retrogression to a 19th century pragmatist’s count of the “scientific” in the 1980s, the Nagel-Hempel bubble was fairly well spent. We can’t get into details here. But we can mention to things, artifacts left over from Rosenberg’s positivist-empiricist/Mill, Carnap & Hume heritage that have left an lasting impact — a continuing commitment to Hume’s associationist psychology image of “causation” and a concomitant metaphysical commitment elemental, simple law-governed categories (such as those know by physics) as the only cognitively significant “entities” of meaning in existence. In short, Rosenberg doesn’t brook the existence of beliefs and desires or any other human categories, the very stuff of economic choice and human learning.
This will be significant when I address in greater detail Hayek’s account of learning in the context of changing local conditions and relative prices as the central explanatory causal mechanism in economic science, a causal mechanism highlighted in part in contrast the to purely logical and non-causal formal construct of the general equilibrium construction.
Rosenberg correctly points out that scientists in Darwinian biology grapple with conceptual problems of the most difficult order, many such problems of which they willingly acknowledge have most fruitfully been clarified and unknotted by professional “philosophers of science” such as David Hull.
Note well that the most difficult conceptual problem in economics — which remains essentially contested — is the explanatory strategy of that science itself. And most significantly, note this: the explanatory strategies adopted by economists have been determinatively formed by images of “science” derived directly from the philosophical tradition and from rumors of nature of “science” filtered through the picture of “knoweldge” and “science” derived from the philosophical tradition. As Hayek, Popper, Kuhn, Wittgenstein and many others emphasize, this image of “science” and “knowledge” is radically false, a mythology of “science” and “knowledge”.
So, at the core of the “conceptual” problems in economic “science” is a contested and deeply problematic inheritance from the tradition of philosophy and especially 19th and 20th century philosophy of science, a tradition that both “mainstream” economic science and Rosenberg share — and one opposed in various ways by figures such as Hayek, Kuhn, Wittgenstein, and even Popper, among others.
So philosophy and rival research problems in the philosophy of science are inherently implicated in the core “conceptual” problems in economics. This is essentially contested territory and the “avoidance” of philosophy or philosophy of science here is a great burying of one’s head in the sand, concomitant with an appreciable ignorance of the history of the foundational role of assumptions about “science” and “knowledge” in the historical unfolding of economic science.
Hayek’s account of the role of the equilibrium construct in the explanatory strategy of economics can be found in the second half of this paper.
“If Economics Isn’t Science, What is It?” by Alexander Rosenberg. Reprinted as Chapter 42 in Readings in the Philosophy of Social Science, edited my Michael Martin, MIT Press, 1994.