On Thursday this week I made the trip to Oxford to see James Ladyman deliver his talk, The Bankruptcy of Analytic Metaphysics. And most entertaining it was too.
The type of methodology which I take James to be attacking, is nicely defined in A Companion to Metaphysics, (Jaegwon Kim and Ernest Sosa (eds.), Blackwell, 1995). In Felicia Ackerman's entry on 'analysis', we are asked to "consider the following proposition.
(1) To be an instance of knowledge is to be an instance of justified true belief not essentially grounded in any falsehood.
(1) exemplifies a central sort of philosophical analysis. Analyses of this sort can be characterized as follows:
(a) The analysans and analysandum are necessarily coexistensive, i.e every instance of one is an instance of the other.
(b) The analysans and analysandum are knowable a priori to be coextensive.
(c) The analysandum is simpler than the analysans...
(d) The analysans does not have the analysandum as a constituent.
(e) A proposition that gives a correct analysis can be justified by the philosophical example-and-counter-example method. i.e. by generalizing from intuitions about the correct answers to questions about a varied and wide-ranging series of simple described hypothetical test cases, such as 'If such-and-such were the case, would you call this a case of knowledge?' Thus, such an analysis is a philosophical discovery, rather than something that must be obvious to ordinary users of the terms in question."
But what, exactly, is the criterion to be applied in these test cases? These are not empirical tests, where we can compare the predictions of theory with the results of measurement and observation. Neither are these tests comparable to the tests devised by mathematicians, to support or reject a mathematical hypothesis. In analytic metaphysics, an attempt is being made to define the meaning of one of the terms of discourse (in the example given here, the term is 'knowledge'); in the mathematical case, all the terms of discourse have been stipulatively defined at the outset.
As Ladyman has emphasized, the problem with this methodology, is that it ultimately appeals to intuition, which is not only culturally dependent, but varies from one individual to another within a culture.
Ackerman acknowledges that "It can...be objected that it is virtually impossible to produce an example of an analysis that is both philosophically interesting and generally accepted as true. But virtually all propositions philosophers put forth suffer from this problem...The hypothetical example-and counterexample method the sort of analysis (1) exemplifies is fundamental in philosophical enquiry, even if philosophers cannot reach agreement on analyses."
It seems to be acknowledged, then, that the results of relying upon intuition are inconsistent. If the results of a methodology are inconsistent, then, in most disciplines, that entails that the methodology is unreliable, which, in most cases, is a sufficient condition for the methodology to be rejected as a deficient methodology. Apparently, however, "all propositions philosophers put forth suffer from this problem," so the methodology continues to be employed in metaphysics, and, for that matter, in epistemology too. Remarkable.