On the demise of the PSR and The Nature of Analytical Philosophy

One attractive story about the demise of the Principle Sufficient Reason (PSR) is that it was discarded in the founding of analytical philosophy together with the heritage of British Idealism and related polemics against Spinoza (and Bergson) by Bertrand Russell. When Russell was given the option, accept (a) Bradley's Regress or (b) the PSR, he chose neither; and he opted, instead, for (c) submission to scientific fact: "The scientific philosophy, therefore...aims only at understanding the world... without being turned aside from that submission to fact which is the essence of the scientific temper." (On Scientific Method In Philosophy [recall my discussion and Jeff Bell.] If the to-be-explained-facts are brute, then it is possible that even if they can be fully captured by integrated into a theory/model (etc.) some arbitrariness is inevitable (in, say, initial conditions). One might even think that this stance is (informally) justified by the "principle of indifference" that accompanies the embrace of a classical probability theory in one's inductive logic (see, Carnap).

Such a story is also attractive because it allows one to explain why after a century of analytical development, a thoughtful philosopher like Thomas Nagel explicitly recoils from "inexplicable...brute" facts, (Mind & Cosmos, 45) and -- exasperating the scientific naturalists among us -- turns to the PSR-lite (recall). Thus, one might think there is a fundamental opposition between scientific philosophy (broadly conceived) and the PSR. (One might even think, with a glance toward that other, Cassirer, that the contrast between "scientific empiricism" and the "principle of reason" is a natural one.) While not all of analytical philosophy is a species of scientific philosophy (we have the common sense brigade, too, of course), much is.

Michael Della Rocca alerted me to the surprising fact that in a crucial step in his well-known, "Carnap on Logical Truth," Quine embraces the the PSR in order to reject Carnap's analytic/synthetic distinction (here). So, the grand-daddy of our naturalists sees no essential opposition between  the PSR and genuine scientific philosophy. This point stands even if one claims that Quine only embraces the PSR for dialectical purposes; for if that is right, he must have felt confident that Carnap would have accepted the PSR in his philosophical methodology.  Indeed, if we turn to Carnap we find:

it is also generally agreed that this determination of extension involves uncertainty and possible error. But since this holds for all concepts of empirical science, nobody regards this fact as a sufficient reason for rejecting the concepts of the theory of extension. The sources of uncertainty are chiefly the following: first, the linguist's acceptance of the result that a given thing is denoted by 'Hund' for Karl may be erroneous, e.g., due to a misunderstanding or a factual erroro of Karl's and, second, the generalization of things which he has not tested suffers, of course, from the uncertainty of all inductive infererence. (Carnap, "Meaning and Synonymy in Natural Languages" (36.)

Carnap's "Meaning and Synonymy" was submitted December 1954. Quine started writing "Carnap on Logical Truth"  "early in 1954" as a contribution to the Carnap Schilpp volume. (I'll leave it to scholars to the figure out how much Quine knew and when.)

Either way, Carnap appeals to the PSR as a proper principle in the develoopment of the theory of extension, which is a proper subject within scientific philosophy. And, in fact, in the passage quoted from Carnap we can find a kind of reciprocal relation between science, where ultimately the principle of indifference rules, and scientific philosophy, where the PSR plays a legitimate role. Note, as an aside, that when a generation later, David Lewis revives Karl in "Radical Interpretation" he locates his decision theoretical project in a "common sense theory of persons," not science (or scientific philosophy).

So, at one point in the middle of twentieth century, leading scientific  philosophers acknowledge that the PSR plays a non-trivial role in scientific philosophy. And, in fact, once one starts fishing one finds it more frequently (e.g., this quote from Reichenbach in Coffa's marvelous book.) This made me wonder, how did the PSR maintain some respectability in the Berlin/Vienna circles?

Cassirer's presentation in "Einstein's Theory of Relavity" of Einstein, the exemplary philosopher-scientist, might help explain this. For he interprets Einstein as gesturing toward an epistemological demand on the form laws of nature can take. In particular, these have to be

independent of the particularity of our empirical measurements of the special choice the four variables...which express the space and time parameters. In this sense, one could conceive the principle of the universal theory of relativity, that the universal laws of nature are not changed in form by arbitrary changes of the space-time variables, as an analytic assertion; as an explanation of what is meant by a "universal" law of nature.--Cassirer, "Einstein's Theory of Relativity," (383-4)

Now, Cassirer respects the fact that philosophy and science have two different tasks. I am assuming that he thinks that it is a philosophical "demand which we make of the form of natural laws." (377; this is so even if the very possibility of this presupposes a historical development of a science.) One striking feature of Cassirer's understanding of science is that at a given time T, scientists may not fully understand the meaning of their own statements (e.g., 364 on Galileo). So, the PSR is a central feature of Cassirer's version of scientific philosophy and the conception of scientific objectivity that epistemology helps secure.

So, what do I take from this little survey? First, the PSR and scientific philosophy are not necessary opposed. Russell might have thought so, but this is a contingent fact about Russell (and, as I have speculated this is due to Boole). Second, there is a task to articulate the proper conception between scientific philosophy and the PSR for us. Third, I leave with a question: why was the PSR not taught as part of the tool-kit of analytical philosophy in, say, the second half of the twentieth century? How come its presence in the Quine-Carnap debate did not keep it alive?