I have once again revised one of my papers which has been looking for a home for quite some time. It’s called ‘Truthmaking and Realism’, and it’s getting longer and longer: now at 11k words. The motivation comes from recent literature concerning truthmaking, especially the OUP book Truthmakers: The Contemporary Debate, edited by Helen Beebee and Julian Dodd (2005). Several authors in the volume suggest that truthmaking is not compatible just with realism, but also with pragmatism and idealism, and thus does not help in defending realism in general. I take this point and suggest that in fact the wider applicability of the truthmaker principle only strengthens the realist’s case, for all that is needed is a plausible way to account for our realist intuitions concerning truth.

To defend this conclusion, I take one of the most influential critiques of realism, which has been advocated by Hilary Putnam, Michael Dummett, and Nelson Goodman; they all share some basic assumptions about realism and its problems. The objection is essentially that realism cannot account for truth. There is an important background assumption here, which is that realism is committed to the correspondence theory of truth. Putnam’s famous model-theoretic argument challenges the correspondence theory by claiming that there will be infinitely many correspondence relations between words and things, and hence indeterminacy ensues as we cannot pick out the intended correspondence. Michael Devitt, in his Realism and Truth (1997), has forcefully argued against the connection between realism and truth, and I take his point, but my argument goes roughly as follows:

1. Let us assume that the Putnam-Dummett-Goodman objection against realism combined with the correspondence theory holds.
2. This does not mean that realism automatically fails, because it is independent of the correspondence theory, as Michael Devitt has argued.
3. We still need an account of truth which is compatible with realism and does not succumb to the Putnam-Dummett-Goodman objection.
4. Truthmaking is such an account of truth, as it is compatible across ontologies, including realism.
5. The combination of realism and truthmaking can stand against the Putnam-Dummett-Goodman objection.

The part that I have revised mostly concerns the Putnam-Dummett-Goodman critique. I’ve added several passages that support my analysis. I still have some reservations about Dummett’s role in all this, because I found passages which suggest that he doesn’t consider realism to be committed to the correspondence theory, such as the following:

The correspondence theory of truth is often claimed as essential to realism. This is evidently false, since Frege was undoubtedly a realist but rejected the correspondence theory. The correspondence theory is also often confused with a truth-conditional meaning-theory, which is the natural extension of the classical two-valued semantic theory that we have taken as characteristic of realism. A properly constructed meaning-theory rightly seeks to characterise the concepts of truth and meaning simultaneously, whereas the correspondence theory took meaning as already given. It is an analogous mistake to regard the principle that, if a statement is true, there must be something in virtue of which it is true, is peculiar to realism. On the contrary, it is a regulative principle which all must accept. (Dummett, The Logical Basis of Metaphysics, 1991: 331).

Still, Dummett here endorses a ‘regulative principle’, which bares remarkable similarity to the truthmaker principle, and this is really all I need for my argument: if the Putnam-Dummett-Goodman critique does not undermine truthmaking, and we have some independent reasons to think that truthmaking is a good starting point for a theory of truth, then the Putnam-Dummett-Goodman critique loses its strength.

I will not go into the details of truthmaking here, and in fact I don’t go very deeply into them in the paper either, but my preferred formulation of the truthmaker principle is as follows:

(TM*) Necessarily, in all possible worlds where [p] is true, there exists something alpha that makes [p] true.

The angled brackets indicate a proposition. The idea is that this principle is compatible across ontologies, but still manages to capture the core idea and plausibility of truthmaking. See the paper for details, any comments are welcome.