A little while back I tackled the perennial question of whether, and in what sense, philosophy makes progress. But that was by means of a fictional dialogue between two robots, part of my “5-minute Philosopher” series, and it’s time to revisit the topic. The occasion has been provided by a lively meetup discussion I facilitated a few weeks ago, based on an article by Toni Vogel Carey that appeared in Philosophy Now magazine.
Carey sets up the discussion by arguing that philosophy stands somewhere between science and the arts, where the first one is the common paragon of a cumulatively progressive enterprise, while within the realm of the latter the whole idea of progress appears to be ridiculous. Although there is much that I agree with in Carey’s article, this set-up strikes me as questionable, particularly because the author counts mathematics as a science. Math is certainly useful to science (and so is logic and, sometimes, even art!), but it ain’t the same thing as science. The latter is concerned with empirically based hypothesis testing, while math makes progress more like logic (a branch of philosophy!), i.e. by a deductive exploration of the consequences of sets of axioms (in logic and philosophy these are called assumptions). So math and logic represent fields clearly characterized by cumulative progress which are not science, thereby undermining the idea that science is the paragon for progressive intellectual enterprises.
Moreover, some of my fellow meetupers even questioned the idea that art doesn’t progress. Yes, as Nobel biologist Francois Jacob (cited by Carey) said, “Beethoven did not surpass Bach in the way that Einstein surpassed Newton,” but the key qualification here is in the (same) way. Beethoven explored ways of composing hitherto unknown to musicians, which has to count as progress in a meaningful (though obviously not scientific) sense of the term. I pointed out during that evening’s discussion that the invention of perspective in Renaissance painting also was an unquestionable case of progress in art, as it made possible painting in ways that were simply not available before. I’m sure other examples can be easily found, especially by historians of music and art.
The heart of Carey’s article, however, concerns three general types of progress in philosophy, each accompanied by an example. The first one is what the author refers to as “progress as destruction.” A lot of what goes on in philosophical research is showing that someone else got it wrong, thereby moving the debate onto higher ground in logical space, so to speak. Carey’s example is Edmund Gettier’s famous demonstration that Plato was wrong when he defined knowledge as “justified true belief.” Gettier did this in a very short paper, using counterexamples. The one Carey provides is actually clearer than the one originally presented by Gettier. Imagine you were watching the final of the US Open a few years back and saw John McEnroe win the match point against Jimmy Connors. Assume further that it is indeed true that McEnroe won the Open that year. Apparently, you have a belief that is both true (McEnroe did win) and justified (you saw the final play). But it turns out that — because of a technical glitch — you actually saw a replay of a similar match point that had allowed McEnroe to beat Connors the year before! Gettier would argue that you have formed a belief that is both true and justified, and yet does not amount to knowledge. Now, put away the discussion of how one could fix Plato’s definition (no one has succeeded so far), because we need to proceed to Carey’s second type of philosophical progress.
This is progress understood as clarification, the sort of thing that Wittgenstein (himself not exactly a shining example of clarity) was presumably thinking of when he said that “Philosophy is a battle against the bewitchment of our intelligence by means of language.” The idea is that philosophers understand certain issues better when they can analytically parse distinct meanings or applications of given concepts. Carey’s example is John Rawls’s analysis of rules within the context of rule- (as opposed to act-) utilitarianism. Rawls distinguished “summary” and “practice” concepts of rules, where the first one works as a heuristic that summarizes past decisions, while the latter examines particular cases of application of a given rule. Without getting into details, Rawls’ approach helped to make sense of the advantages of rule-utilitarianism over act-utilitarianism, at the same time that it also made clear that rule-utilitarianism is barely utilitarianism at all, and falls uncomfortably close to its chief rival, deontology (i.e., rule-based ethics).
The third and last situation considered by Carey is “progress as doubt,” in which philosophers provide a needed counter to over-enthusiastic practitioners of their own and of other disciplines (e.g., science), by pointing out just how much we really don’t know. Here David Hume’s famous problem of induction comes to mind. Hume argued very effectively that induction — on which much everyday reasoning and especially scientific inference are based — cannot be logically justified on independent grounds. (If you think you can get out of this by arguing something along the lines of “induction works” think again: that would be invoking inductive reasoning to support inductive reasoning, and you’d be open to one of the worst charges in philosophical reasoning, that of circularity.) One cannot avoid but think of Socrates, and of the Delphi Oracle’s statement that he was the wisest man in all of Greece, apparently on the basis that he knew that he didn’t know much.
There are certainly other examples one could line up following Carey’s approach. Quine’s criticism of the previously universally accepted distinction between synthetic and analytic statements; Popper’s proposal that scientific hypotheses have to be falsifiable, followed by a Duhem-Quine inspired argument showing that falsification doesn’t work; the increasing sophistication of different versions of utilitarian ethics (from Bentham to Mill to Singer); the various moves and counter-moves in the debate in philosophy of science between realists and anti-realists; and so on.
What all of these modes and examples of progress in philosophy have in common is that they use analysis to parse and explore the logical space in which philosophical discourse exists. One begins with a given set of assumptions and works out their implications, until someone points out a problem with some of those implications which requires either the addition of other postulates or the abandonment of the initial one and their replacement by another set that may work out better. In this sense, philosophical analysis, again, is much more similar to mathematics than to science, and the discipline of logic represents a great example of it, both because it is a branch of philosophy that has clearly made progress, and because it can be said to actually include mathematics, at least in the sense that math is also about the application of deductive reasoning to uncover the properties of systems of axioms. That said, of course, I do not expect my colleagues in the math department to move in with us, though they would certainly be welcome...