ABSTRACT
We come, in this chapter, to a consideration of truth and we shall treat this subject in the same way as we have those in Part I, by first giving a brief picture of different views of truth, and then developing the view that I myself will maintain.We could say with Russell (1940): Truth is a property of beliefs, and derivately of sentences which express beliefs. Truth consists in a certain relation between a belief and one or more facts other than the beliefs. or we could sensibly say with G. E. Moore that what we mean by truth is self-evident and everyone knows what it is even if we find it difficult to put into words. If I hold out my hand and say “this is my hand”, no one seriously doubts that I am making a true statement, although we are hesitant about using words like ‘certain’ (Wittgenstein, 1969). The difficulties arise only over the limiting cases, in much the same way as illusions and hallucinations create difficulties for our theories of perception.What we can say is that truth is a correspondence between a proposition and reality, so that if I say “France beat New Zealand at rugby football” this is a statement of fact that is not in
itself a proposition, but is, as it were, ‘a token of a type5, but if it corresponds to the fact that they did (i.e. France did beat New Zealand) then the statement is true. We can, of course, add any amount of detail about the score and the names of the players, the precise date of the match and so on and so forth, but the simple highly abstracted statement is still true or false according to whether there is this correspondence or not between the statement. This is what Price (1953) calls the physicalist version of the Correspondence Theory.We are assuming, of course, that truth or falsehood can only apply to meaningful indicative statements, and this is by no means all of language. This means that truth presupposes meaning and the question “What do you mean?” always precedes the question “How do you know?”. Furthermore, we shall not (cannot) talk of the truth of commands, opinions and the like. Our commitment to critical realism (Chapters 3 and 4) entails in effect a commitment to a correspondence theory of truth.The Coherence Theory of truth depends at least upon the consistency that statements have with each other, or indeed the coherence between the beliefs we hold could be said to be a guarantee or a warranty of their truth. But this is not so. It is patent that a thoroughly coherent and consistent story can be composed of falsehoods. So that, while we accept Peirce’s view that there is a convergence towards truth in the beliefs held by a bulk of people, this cannot be a sole or sufficient criterion of truth, nor indeed was normally intended to be.Coherence, in fact, provides evidence for the truth of beliefs and their statement in the same way as empirical statements can be more or less confirmed by evidence. We can, in practice, only know the degree to which empirical statements are confirmed: we cannot know that they are true or false. In practice, we most often will “know”, but it is quite obvious that to push that knowledge too far could be risky. The measure of the risk, if action or decision is entailed, will help us as to whether to act or decide.If we say that correspondence is a correct depiction of what we mean by the word ‘truth’, more often than not we can easily accept the semantic theory (or rule) of truth of the form “ ‘the cat
is black’ is true (a true statement) if and only if the cat is black” (Tarski, 1944). A typical reaction to this from non-professional philosophers is the derisive laughter normally reserved for the obvious. But then what we mean by truth is, for the most part, obvious.In fact, the full implications of the apparently obvious are less than obvious since the Semantic Theory emphasises the importance of distinguishing language from “facts” and makes clear that it is the correspondence between the two which is paramount.Similarly, we find it easy to incorporate the Pragmatic Theory into our system of things by accepting that if we suppose certain statements to be true and the consequences of that supposition turn out to be highly predictive, then for all practical purposes they (the original statements) can be thought of as true, and of course may well be. This view bears a somewhat similar relation to the Correspondence Theory as does the Coherence Theory. We shall return later to consider each of these views in rather more detail.As we have already said, the very fact that we have adopted a critical Realist view of ontology suggests that we advocate a correspondence theory of truth, but we will stand back a little from the problem and consider coherence, especially since Wittgenstein in the Tractatus thought of basic statements as corresponding to reality, while the derived complex statements depended upon their consistency with the basic statements. This perhaps allows the notion of coherence (at least as consistency) a foot in the door, for Wittgenstein makes truth a matter of both correspondence and coherence.Wittgenstein says in the Tractatus: A picture agrees with reality or fails to agree; it is correct or incorrect, true or false. (2.21) If the world had no substance, then whether a proposition had sense would depend on whether another proposition was true. (2.0211) In that case we could not sketch any picture of the world (true or false). (2.0212)
But later (Philosophical Investigations, 1953) he says: It is quite true that, in giving an ostensive definition for instance, we often point to the object named and say the name (my italics). (19e) There are many such examples which demonstrate clearly that he follows Moore’s view in assuming ‘true’, in fact, is well understood. And again, in On Certainty (1969), truth appears in its correspondence role. This no doubt represents a distinction between the everyday use of ‘true’ and the philosopher’s discussion over ‘true’. Let me now change tack and look at truth more in the manner of Carnap and the formalists. THE FRAMEWORK OF TRUTH We shall now consider two approaches to truth: the first is to give an account of how the word ‘truth’ is used in natural language, and this will not be our primary aim, since such an inquiry should be preceded by — or be part of - an empirical investigation as to how the word is actually used and understood by enquiring from a statistical sample of people. This approach may also involve an explication of the concept (and the word) in terms of a theory of behaviour, which we do not aim to carry out here: but the importance of such an investigation should not be overlooked. The more usual philosophical alternative, which is our primary interest, is to lay down rules for the use of the term which is in keeping with our intuitive notions of how the term is used by the majority of people - examples of actual usage can be our rough guide, as we have already shown with a few examples from Wittgenstein.This second rather characteristic procedure can itself be undertaken from at least two different points of view. We can in the first place start from a given language that is acceptable to all (more or less so) and give an account, definition-in-use (contextual definition) or explicit definition of ‘truth’ in a metalanguage of the particular object-language with which one is concerned (we shall be discussing definitions of various kinds in Part III). A well-known solution along these lines has been found by Tarski, as we have already mentioned. His theory (or
rule) of ‘truth’ seems satisfactory for most purposes, and certainly so for those of formalised languages.I would like to make it clear at this point that some confusion can arise over the criteria, definition and meaning of ‘truth’. I would say that the meaning o f‘truth’ is given by the criteria and is embodied, therefore, in the definition. We are here talking (see Chapter 10) of a word-word rather than a word-thing definition. In the case of truth, then, we are thinking of word-thing or even thing-thing definitions; Tarski’s rule can be thought of as word-thing in an obvious sense.We are involved, though, with more than language. We are also involved with an investigation from an ontological viewpoint that assumes that ‘truth’ involves more than a linguistic rule and refers also to matters of knowledge fact or belief). This last statement implies that we should take another look at the theories of correspondence and coherence, and that we should also review the status of the Pragmatic Theory of truth.The term ‘ontology’ or ‘ontological commitment’ is used here to stand simply for “what there is”. We think of this as being involved in the conceptual framework of the language chosen - we are committed to an ontology which implies the existence of physical objects. It is through our bound variables that we run into such ontological commitments, whether they be Phenomenalist, Realist or whatever.Before it is possible to take up these points it is essential that we should turn to the well-known distinction made by Carnap (1952) between ‘truth’ and ‘confirmation’. ‘Truth’, he says, is a time-independent term. It applies to statements (we shall use the near-neutral term ‘statement’ to stand for either ‘sentence’ or ‘proposition’) in space-time and allows us to categorise such statements as are amenable to factual interpretations as either true or false. This is clearly an explication of what is supposed to be commonly meant by ‘truth’. “Confirmation”, which is time-dependent, refers on the other hand, not to the truth or otherwise of a state of affairs as given by a statement, but to the extent of our knowledge of that truth or falsehood. Thus we have to distinguish between a “state of affairs’ and our knowledge of that state. Carnap’s distinction rests on examples such as:
1) “The substance in this vessel is alcohol.”2) “X believes the substance in this vessel is alcohol.”3) “X knows that the substance in this vessel is alcohol.”
