Theories of Truth

 shall briefly summarize the main philosophical theories of truth:

• Correspondence theories (Russell, 1912; Tarski, 1944; Popper, 1959; Wittgenstein, 1974) are the main and most obvious view of truth. They hold that truth (and falsity) is applied to propositions, depending on whether the proposition corresponds to the way the world is. It thus applies to the relationship between a proposition and the states of affairs it describes. Problems with this view are: (i) In what sense can a linguistic statement be said to correspond to something quite different – an occurrence in the world? (ii) We cannot directly access the external world so we are only ever comparing experiences and statements with other experiences and statements, so that we can never actually determine if a proposition is, in fact, true. Most other theories stem from the problems in maintaining a correspondence theory.
• Coherence theories (Bradley, 1914; Putnam, 1981; Quine, 1992) stress the extent to which a proposition is consistent with other beliefs, theories and evidence that we have. The more that it fits in with other well-attested ideas, the more we should accept it as true. This approach avoids the need for a direct comparison with ‘reality’. However, it is more concerned with the justification of beliefs rather than their absolute truth. From a Kuhnian (1970) perspective, fitting in with the current paradigm does not make the current paradigm correct. Quine held that coherent systems of beliefs were under-determined by empirical data and thus that no theory could ever ultimately be verified or falsified.
• Pragmatic theories (Peirce, 1878; James, 1976; Rorty, 1982) hold that truth is best seen in terms of how useful or practical a theory is – that which best solves a problem is the best theory. A version of this is instrumentalism, which holds that a theory is simply an instrument for making predictions, and has no necessary connection to truth at all. This also leads into consensus theories. An obvious argument against this view is that a true theory is likely to be most useful and powerful³ and therefore should be an important component of a useful theory.
• Consensus or discursive theories (Habermas, 1978) accept that truth is that which results from a process of enquiry resulting in a consensus among those most fully informed – in the case of science, scientists. At one level, we can see that this must be the case if we accept with CR the impossibility of proving correspondence truth. But, often today's accepted truth is tomorrow's discarded theory and so this does not guarantee truth. See the discussion below about Habermas's more recent views.
• Redundancy and deflationary theories (Ramsey, 1927; Frege, 1952; Horwich, 1991) argue that the whole concept of truth is actually redundant. If we say ‘it is true that snow is white’ we are saying no more than that ‘snow is white’, the two propositions will always have the same truth-values and are therefore equivalent. This seems to me largely a linguistic move as it does not touch upon the question of how we might know or believe that a proposition is actually the case.
• Performative theories (Strawson, 1950) also deal with the linguistic use of the term. The suggestion here is that by saying ‘p is true’ we are not so much commenting on the truth of the proposition as such but on our willingness or intention of accepting it as true and commending it to someone else. Again, this just seems to ignore large areas of the question of truth.

Mingers