|
In my review of Michael Beaney’s entry on analysis for the online Stanford Encyclopedia of Philosophy, I made several statements to which he takes exception. His main criticism is that my review fails to appreciate the rich varieties of analysis that he was trying to show exist in philosophy. Instead, it focuses on only three types discussed by him – regressive, decompositional and interpretive analysis – and ignores the other kinds he mentioned and also ignores the complex interrelations between these various methods. I am glad that Beaney has taken the opportunity to stress this point in his reply to my review. I agree that it is much more interesting to look at the various periods in the history of philosophy as a rich complex of methods rather than as consisting of a few methods that can be categorized as belonging to one of three types, and it is true that he mentions and describes many more than these three kinds of analysis. Still, it is decompositional, regressive or interpretive analysis that Beaney mainly discusses, and he most often says of some method of analysis, after describing it, that it is like one or more of these three major types, which is why I focused on them. What other types of analysis does he mention? One that he discusses throughout the text is “reductive” analysis. In it, a concept is said to be “reduced” to others when one can eliminate it in favor of the other concepts, for example, in one’s description of the world, in which case one has discovered a metaphysical fact about the world. Beaney principally describes reductive analysis as a kind of interpretive analysis that interprets talk using the concept into talk without it, though I suppose decompositional analyses can be equally reductive. Beaney discusses reductive analysis at length in the case of Gilbert Ryle and other Oxford analysts, making the point that Ryle shifted from this principal method of analytic philosophy to a kind of non-reductive interpretive analysis that Beaney calls “connective” analysis. Where one cannot eliminate a term without circularity, so that it is in some metaphysical sense irreducible, one can still (circularly) clarify it’s meaning by interpretating it in terms of other concepts that can only be defined in terms of it. Such an interpretive analysis is a connective one, one that shows the logical connections between these basic irreducible concepts. Though I did not quite gather what claims are made for such analysis, that is, what its significance is supposed to be (are these the “true meanings” of the analyzed concepts or are they something more arbitrary), Beaney makes the intriguing suggestion that “connective analysis would seem to be particularly appropriate … in the case of analysis itself” (that is, I think, in analyzing analysis itself). It needs to be emphasized however that my point in discussing the three kinds of analysis I saw Beaney spending most of his time on was not to say that they are inadequate to describe the varieties of analysis in philosophy and that we need a richer taxonomy than he provides, but that there are certain questions about analysis which Beaney’s explication of even these three types does not answer for me. Nor do I take this as an inadequacy of Beaney’s discussion, since his purposes are not to answer my questions, but his own. Beaney refers to some of my concerns when he says that I am right in saying that he needs to address the issue of the apriority of these and other methods of analysis. My point, though, is more general than this. What puzzles me whenever I hear people talk of analysis or see such ideas in print is how exactly the analysis is supposed to work. What are the specific steps that one takes in each kind of analysis, and most important, how does one justify each of these steps? (It would also be interesting to learn how it is thought that we psychologically move from step to step, how we are thought to discover the various steps in an analysis.) When asking how a step in an analysis is justified, the question of whether the justification is supposed to be a priori or a posteriori does arise, but so do other questions such as what metaphysical assumptions are being made in each case, and since it is usually concepts that are being analyzed, what theories of meaning are being presupposed. (These questions might also arise when trying to say how the various steps of an analysis are arrived at, that is, discovered.) But these are my questions. This being the case, perhaps it is up to me and no one else to answer them. Beaney has done an impressive job of answering his own questions, one that that I think will inspire others to try to push the subject even further. To move back to Beaney’s criticisms of my review, he points out that right at the start of it I attribute to him a general characterization of analysis as being decompositional when that is not what he says, and I am chagrined to see that I do make this error. His most general characterization of analysis comes in the second sentence of the piece and is that “in its broadest sense, [analysis] might be defined as disclosing or working back to what is more fundamental by means of which something can be explained (which is often then exhibited in a corresponding process of synthesis)….” As he notes, he then goes on to emphasize that decompositional analysis (breaking a concept down into more simple parts) is not the only sort of analysis that philosophers have practiced, and is arguably not the most important one. Beaney also points out that I suggest that he tries to give a single characterization of analytic philosophy and must then immediately admit Moore as an exception. I did erroneously suggest this and I withdraw the suggestion. Beaney emphasizes that there are several major strands of analysis to be found in analytic philosophy and that Moore’s is one of them. (He also points out that the “knowledge is or isn’t justified true belief” discussion in analytic philosophy is another instance of decompositional analysis.) A final quibble with Beaney though. He says in his reply to me that it is not his aim to pigeonhole philosophers into various categories or impose a rigid taxonomy. I hope this is just a matter of emphasis and that he will not wholly dismiss taxonomic tasks in his further work on the subject, and I suspect that he won’t, for he also says that “of course, some kind of conceptual framework must be developed to elucidate the various forms of analysis and their interconnections”. For my part, I do not think that any historical period can be accurately discussed or even clearly thought about without a good taxonomy and genealogy of its ideas, and the game of taking someone’s taxonomy (even one’s own) and trying to refine or modify it or elaborate on it is an important and probably essential way of moving the understanding of a period forward. Philosophy & Religious Studies Dept.Youngstown State University Youngstown, OH 44555 jcongley@ysu.edu |