Haack on formal philosophy

Posted on September 26, 2008


I have something of a love-hate relationship with formal methods. On the one hand, having done my share of logic courses, I have a deep respect for what a set of good formal tools allow us to do. On the other hand, I find that all too often people who work on such approaches are guilty of over-reach, of thinking that formal approaches can provide all the answers. I think an awareness of this over-reach is, at least in part, behind post-modernist criticisms. Of course, most of criticisms are way off, having very little appreciation for what they are criticising. And they certainly do not bring us any closer to the kind of half-way point that I would like to find solid footing at. That is why I was so pleased to read Susan Haack’s article on formal philosophy. Haack has done a lot of work on logic and was asked by Vincent Hendricks and John Symons to write answers to several of their questions. Her replies are often very enlightening, even if that paper is somewhat uneven due to its Q&A format. Here’s one point Haack makes that I think particularly needs to be remembered:

Reading in the history of logic taught me to distinguish a broader, Aristotelian conception of the subject, and a more modern, Fregean one. In the broader sense, of logic as the theory of whatever is good in the way of reasoning, its scope would include not only formal, systematic representations of deductively valid arguments and meta-theoretical results about such systems, but also theories of non-deductive reasoning and theories of the term, of propositions, truth, etc.. These days, however — rather as, thanks to the success of particular brands, “Xerox” has become a generic word for photocopying, and “Kleenex” for tissues — “logic” seems usually to refer, quite narrowly, to syntactically characterizable systems of valid reasoning, of which the Frege-Peirce-Russell propositional and predicate calculi would be the paradigm. Dewey’s Logic: The Theory of Inquiry, for example, though it surely is logic in the older, broader sense, bears little resemblance to logic in the newer, narrower understanding.

The distinction is an important one if one wants, as I do, to be able to say that “syntactically characterizable systems” do not actually characterise rational thinking, without falling into the oceans of postmodernist vagueness, and while retaining as understanding for why such systems are as useful as they are.