274. Is pragmatism conventional?  

It has been claimed (e.g by Richard Rorty) that pragmatism is conventional: ‘treating conventionally accepted norms as foundations’[i]. I am a pragmatist and yet I disagree, up to a point. If pragmatism were conventional, it would be inherently conservative, and I propose that pragmatism can support novelty.

What is conventional? I propose that it can be rendered as operating within an established language game. Certain terms, meanings and ‘rules of the game’ are taken for granted. In science, it could, I proposed in the preceding item in this blog, be rendered as preserving the ‘core’ of a ‘ research programme’, in the terms presented by Imre Lakatos: fundamental theoretical and methodological principles that are not susceptible to falsification. Empirical anomalies are to be dealt with by means of alterations in a ‘protective belt’ of auxiliary assumptions.

In my view, as I argued before (in item 264 in this blog), something is to be accepted as ‘warranted’ when it ‘works’, logically, empirically and practically. I propose that this does not necessarily require fit in some existing language game, and hence can escape conventionalism in that sense. I grant that it does presuppose some agreement on criteria concerning logic, empirical testing, and practical success across language games. There is no rock bottom for truth beyond any and all perspectives, but we may not stand empty handed in trying to step out of a language game, or a research programme, into a wider, more generic one. I am not claiming that this is always possible, and that there is some ultimate, authoritative language game that can decide universal legitimacy.

There is overlap of at least some terms, principles, assumptions, perspectives, between language games, even if they are in different languages (English and French, say). If terms are shared, they are not likely to have identical meaning, since meanings depend on relations between terms in the game, but, I propose, they are likely to have some family resemblance if they are used across games.

To be specific, let me expand a bit on a project for a radical transformation of economic science, which I mentioned in the preceding item in this blog. That is based on radically different perspectives on human conduct, ethics, scientific conduct, and the notion of uncertainty. Many economists reject this out of hand. However, I do employ some established concepts from economics, (such as ‘transaction costs’), though twisting and extending them a bit, and I refer to phenomena that economists might acknowledge (though they look differently on their relevance for theory). My ambition is to show that alternative theory explains certain facts better if only one accepts them as relevant. That ambition may fail, but it is not necessarily hopeless.

Can a pragmatist offer rigourous arguments? Richard Rorty said he/she cannot because rigour requires unshakeable foundations, which the pragmatist does not accept. Again I disagree. He confused rigour with certainty. One can have rigorous arguments on uncertain foundations. Take mathematics. It is rigorous on the basis of uncertain, merely assumed axioms. The grounds for rigour may shift, but they are still there for some time or in some area. Euclidean geometry was supplemented by other geometries. It applies on a plane but not on a sphere.

I agree (with Rorty) that rigorous argument requires a shared language game, terms with shared meanings, shared assumptions, shared grammar or method (rules of the game), and shared explanatory goals.

Compare this with Thomas Kuhn’s notion of a ‘ paradigm shift’ involved in breaking the rules of a game, stepping out of the game, resulting in ‘ incommensurablity’, an impossibility of rigorous argumentation between games.

But, as I suggested, one may still have the benefit of a wider, roomier, more general game. A different ball game is still a ball game. Parts of argumentation may show a family resemblance between language games. I do think that discussion between language games involves differences of meaning and intention, and therefore is always imperfect tinkering, and often does fail. Moving between games is more like literary narrative than like rigorous scientific discourse. That may be rejected as unscientific, and then debate is indeed hopeless.

What games are there in philosophy? I take this question also from Richard Rorty. One game is to take philosophy as ‘transcendental’, reflecting on the conditions under which some theory or practice (concerning truth, reality, or morality) is possible. But what are the conditions for such conditions to be possible? It yields an infinite regress of conditions for conditions. The underlying intuition is that there are, must be, independent, fixed principles to build on. 

Another game, going against that intuition, is that of anti-essentialism, anti-foundationalism, as in pragmatism. Think of philosophers Peirce, Dewey, Nietzsche, Wittgenstein, Quine, Derrida, and Rorty.

Are these two games incommensurable, with no recourse to sensible debate? One may think of the supposed rift between analytic and continental philosophy. However, they still have things in common, such as the themes of knowledge and morality, even if they differ fundamentally in their views on them. Disagreeing on fundamentals, they may still compare implications for science, politics, economics, literature, …. They may even agree that in some cases the other side seems to be making sense. And indeed, some bridging between analytic and continental philosophy does seem to be taking place.        

---

[i] E.g. by Richard Rorty, in an essay on Derrida, in Essays on Heidegger and others, Cambridge U. Press, 1991, p. 119