Over and above good argumentation, philosophy seems to strive after a criterion to settle disputes. Pragmatists provided such a criterion stressing the relevance of practice for philosophy, joined in this by the later Wittgenstein. What we actually do and say cannot but have a spin-off on philosophical reflection, if the latter should be of any worth.

As to the interpretation of the concept of truth, it is a remarkable fact that two of the leading contemporary (neo-)pragmatists – Hilary Putnam and Richard Rorty – subscribe to the so-called deflationary conception of truth (or something very close to this) as a consequence of their appeal to both practice and the thought of the later Wittgenstein. Equally remarkable is the fact that Paul Horwich, who has the great merit of having revived interest in alethic deflationism, has done this by putting himself in the wake of Wittgenstein. The (direct or indirect) appeal to practice represents thus a sort of fil rouge connecting the three philosophers – Horwich, Putnam and Rorty – as far as the notion of truth is concerned. Yet, differences among the three about the actual interpretation of alethic deflationism is evident.

Therefore, questions like the following spontaneously arise: Who offers the best account of our verbal and non-verbal practice? And, consequently: Who offers the most plausible version of alethic deflationism?

The paper will try to find an answer to these questions, tackling the differences among the three philosophers in the first place. To this end, a possible route is to ascertain what they respectively believe about the nature of truth.

In accordance with the usual deflationary view, both Horwich and Rorty think that truth has no nature. This is because, according to them, the concept of truth is “empty” and thus there is no nature to describe. In a nutshell: “no content, no nature”. Moreover, the emptiness of the concept reveals that there is no property of truth one could ascribe to one's favourite bearer of truth (and, a fortiori, there is no normative property). The philosophical analysis of truth reduces itself to the analysis of the uses speakers of a language do of the relevant words – “true” and cognates. Embracing a view made popular by Wittgenstein (this goes for Horwich in particular), we have that it is “use” that places a meaning on words and sentences. In particular, as far as sentences are concerned, use amounts to satisfy assertibility conditions – not truth-conditions. According to this line of reasoning, this is so (i.e., any account of the understanding of sentences should appeal to assertibility conditions) because, being truth banned from the set of the philosopher's explanatory tools, truth-conditions appear useless. Horwich's and (arguably) Rorty's positions are therefore based on a verificationist semantics.

Putnam does not take side with Horwich and Rorty, in this respect. In fact, he is willing to acknowledge a “physiognomy” of truth – if not a real nature (at least, this is what the paper will try to clarify). In fact, reflection on practice makes Putnam realize the centrality of truth for the speaker/hearer of a language. For one thing, our verbal practice reveals that we call a sentence “true” when there is a normative way of establishing that it is true. For another, it reveals that understanding sentences is not only a question of assertibility conditions, but also of fully legitimate truth-conditions (depending on the sentence in question, of course). Both elements make truth appear as something “substantial” – something very different from the idea according to which talk of truth is a mere façon de parler.

Of course, thinking that the predicate “is true” ascribes a normative property to statements departs from the usual understanding of truth as deflationary. The final part of the paper will be devoted to ascertain whether or not it is right to see Putnam's conception of truth as deflationary.