IS ECONOMICS SCIENTIFIC? IS SCIENCE SCIENTIFIC? – Part 1
By Phin Upham
Part 1
ABSTRACT: The usefulness of models that describe the world lies in their simplicity relative to what they model. But simplification entails inaccuracy, so models should be treated as provisional. Nancy Cartwright’s account of science as a modeling exercise, in which fundamental laws hold true only in theory—not in reality, given the complexities of the real world—suggests that Rational Choice Theory (RCT) should not be rejected on the traditional basis of its lack of realism: that, after all, is to be expected of any simulacrum model. But when RCT has been extended to domains, such as politics, in which there is no necessary reason to expect systemic pressures against people who depart too far from the model, RCT is a simulacrum without any particular claim to expressing underlying causal laws. This cautions against the tendency to rest content with models and to treat their assumptions as if they were true. The question of whether economics and other social sciences are really “scientific” is often answered by taking some view of the natural sciences for granted, and then seeing how well the social “sciences” measure up. But an equally illuminating approach might be to see how a prominent challenge to conventional views about natural science affects our understanding of social science.
Critical Review 57 (2005), nos. 1-2. ISSN 0895-3851. www.criticalreview.com
Phin Upham, phin.upham.wpos@wharton.upenn.edu, is a doctoral candidate at the Wharton School, University of Pennsylvania, 2061 Steinberg Hall-Dietrich Hall, Philadelphia,S. IS ECONOMICS SCIENTIFIC?
IS SCIENCE SCIENTIFIC?
I 18 Critical Review Vol. 17, Nos. 1—2
Scientific Laws in a Complex World
At the heart of Nancy Cartwright’s unconventional theory of natural science is the rejection of fundamental laws of nature as true statements about reality. “Nature, tends to a wild profusion,” according to Cartwright (1983, 19), “which our thinking does not wholly confine.” Nature is inherently unpredictable and exceedingly complicated. In her Cogito interviews, Cartwright says that as a result, reality and the results of a scientific model will never mesh perfectly (Pyle 1999, 208). One can construct artificial laboratory conditions carefully, but the resulting match will be asymptotic: the predicted results might approach, but never mirror, reality.
Cartwright contrasts phenomenological laws—laws about appearances—against fundamental laws: laws about the reality behind the appearances. The latter are meant to explain phenomena, not describe them. “Really powerful explanatory laws of the sort found in theoretical physics do not state the truth” (Cartwright 1983, 3), since the only truth to which we have access regards descriptions of phenomena.
Moreover, descriptive accuracy is lost, she argues, when we turn from phenomena to laws meant to explain them. Separating the laws of physical science from truth is counterintuitive in our mental universe, which is steeped in realism. As a layman, one is used to believing Without second thought that forces such as gravity actually exist and that objects on the Earth fall as a result of this force. After all, scientists have been able to use “the laws of gravity” to predict events all across the galaxy with great accuracy. Can such an ubiquitous and seemingly verifiable force be denied validity, Einstein notwithstanding?
Cartwright answers by using the example of a store-bought radiometer, the familiar children’s—science contraption on which, inside a glass—enclosed partial vacuum, are perched several vanes painted white on one side and black on the other, poised to spin when exposed to light. No theory yet devised, she claims, can predict the exact rate of spin when light shines on the device, nor will any theory yet devised completely and adequately explain why it turns. Our theories (fundamental laws) fail to adequately explain observed phenomena. “Realists are inclined to believe that if theoretical laws are false and inaccurate then phenomenological laws are more so. I urge just the reverse” (Cartwright 1983, 3). Cartwright maintains that laws of phenomena are descriptive, accurate, and verifiable, while fundamental laws are not.
It is widely held that if a hypothesis explains a large number of phenomena without falsification, this provides strong evidence that the hypothesis may be considered a law. Cartwright rejects this standard, demanding that any model of explanation must meet the challenge that Bas van Fraassen presents in The Scientific Image. Van Fraassen contends that to link explanatory power with truth, there must be some guarantee that if x explains y and y is true, then x is also true. If this link does not
hold, then x is merely a plausible explanation of y.
Only causal relationships satisfy van Fraassen’s demand. “The general success of [a] theory at producing accurate predictions, or at unifying what before had been disparate, is no help” when dealing with the crucial issue of causal roles, according to Cartwright (1983, 8). Insofar as our belief that x has caused y entitles us also to believe that x is true, “causal reasoning provides good grounds for our belief in theoretical entities.” Thus there is “no inference to the best explanation, only inference to the most likely cause” (ibid., 6).
Cartwright’s approach does not disregard the organizational, predictive, and inspirational usefulness of fundamental laws. She believes however, that different fundamental laws apply to different situations with different purposes. Each lawlike model focuses on different elements of a phenomenon. There is no correct model for every situation. “A theory is successful in its domain” (Cartwright I999, 31). Cartwright stipulates that this conception of laws holds because neither nature nor reality is unified by such laws, contrary to the usual realist assumption.