In a very subtle 1999 paper [that I read in a reading group with Morgan Bennett at UCSB] Nancy Cartwright adopts the stance of a revisionary "empiricist" (333) in which she rejects the stance of "many [other] empiricists" that "causal laws and probabilistic — or, in the limiting case, deterministic— relations are fundamental to nature." (320) She argues with brilliant panache, instead, that "exact sciences when they are most successfully done seems likely to confine them within limited domains." (318; Her examples are from physics and economics.) To put her point a bit too simply: the theoretical content embedded in mathematical models apply to the exact domain of application of these models and no more. In doing so, Cartwright unintentionally recovers an eighteenth century tradition of empiricism (Mandeville, Berkeley, Hume, Diderot, Buffon maybe Adam Smith) with roots in Locke (and Spinoza) that used a version of this strategy in order to limit the authority of newly ascendant Newtonian mathematical physics. (I nod at this tradition here.) Here I ignore the metaphysical doctrine behind Cartwright's approach. An important feature of Cartwright's argument is that she embeds it in a normative project for science: "the central aim of the sciences is to make the world the way it ought to be." (333) For Cartwright the role of science is to lead to actions that will change the world for the better. (Again, the eighteenth century tradition is also heavily focused on making science more useful.) In other words, her empiricism is a species of pragmatism.
In her paper Cartwright casually mentions two of her target empiricists as "Philosophers like C. G. Hempel and Ernst Nagel" (331), but she barely engages with their positions. Now, as regular readers of this blog know, I follow Schlick in treating Hempel as the wrong sort of empiricist who mistakes science for the world (and, thus, is not able to be properly philosophical about it). But Ernest Nagel, the dominant American philosopher of science between 1934 and the 1960s, is not so easily shunted aside. Now Nagel is now best known among philosophers of science as the canonical source for treating intertheoretic reductions. He is also known for introducing so-called bridge-principles into philosophy of science (and this is how Cartwright nods at him in the paper). But here I want to focus on the ironic fact that Nagel, too was a pragmatist-empiricist not so very different from Cartwright. To see this in action [sic] we must turn to an overlooked gem: Nagel's (1963) response to Milton Friedman's then already "well known" (1953) methodology paper. (The fact that Nagel is the philosopher -- among notable economists -- that had been invited to comment on Friedman at the American Economics Association should come as no surprise; his text-book was widely read in economics (see p. 20 of George Stigler's recommended readings in his 1942 textbook The theory of competitive price.))
Now, in the process of defending Friedman's conclusion (but not his argument) Nagel takes on an important economist-critic of Friedman, Tjalling Koopmans (like Friedman and Stigler, a future Nobel Laureate and also the co-founder of econometric macro-modeling). As will be clear when I quote Nagel's criticism of Koopmans, Koopmans defends an earlier version of Cartwright's empiricism. (No surprise, because Koopmans designed his science with policy ends in mind.) I quote Nagel's concluding passage (and then offer a brief comment): "Thus Professor Koopmans argues that if (as Professor Friedman holds) the fact that firms whose behavior diverges from it are not likely to survive is a basis for accepting the hypothesis, "we should postulate that basis itself and not the profit maximization which it implies in certain circumstances." This seems like a recommendation that since a basis for accepting Newtonian gravitational theory is the fact that observed regularities in the motions of the planets are in agreement with various special laws deduced from the theory, we should postulate those regularities rather than the theory-a recommendation that would replace the theory by the empirical evidence for the theory. Such a proposal not only rejects the conception that theories have an explanatory function; it also overlooks the irreplacable role theories have in scientific inquiry in [A] suggesting how empirical generalizations may need to be corrected, as well as in [B] directing and systematizing further empirical research." Ernest Nagel (1963) [I have defended Friedman along similar lines in 2005--mea culpa, I could have known Nagel's piece.].
Now, Cartwright self-consciously rejects such appeal to "explanatory function" (330ff). But in [A] and [B] Nagel, ever the Pragmatist, has put his finger on the research-guiding utility of law-talk that seems to have eluded her. ([A]&[B] echo Newton's fourth rule of reasoning, which has been explored by George Smith in his writings.) So, to put the moral of this story simply: the price we pay when we adopt Cartwright's austere, normative empiricism (which is offered in the spirit of offering "methodologies for the messy world that we inevitably inhabit" (335)) is, if Nagel and Newton are right, a less powerful όργανον for discovery. Or, more ironically, the (religiously-tinged) obsession with explanation serves the needs of research quite well. To put this in terms borrowed from David Lewis, it is a good myth. [Thanks to another UCSB graduate student, Alex Leibowitz, for the reference to Lewis.]
Recent Comments