What is Truth?Unfortunately for the metaphysician, the very question being asked when someone queries 'What is Truth?'is not very clear. Are we, for example, asking for a definition of how people use the word, or perhaps how they should use it? Are we asking what use Truth might be put to? Perhaps we are asserting that there is such a thing, in the 'real world', as true facts, and asking what the nature of these are. If we are to answer the question, we must first decide which one we are asking. The discussions in the literature, however, tend not to focus on any one of these questions, but in fact different authors attempt to answer different formulations. I hope to show that a number of different theories have different advantages, and that few are as diametrically opposed as some would claim, and thus various can be held together, rather than being competing theories. I shall also try to show that any formulation of what Truth is that does not attempt to address how people use the word is misleading, and not answering the question at hand, and will discuss in this light to what answer the question might lead.
Those, for example, offering a Redundancy Theory of Truth seem to be answering a question which is fairly distinct from that which anyone else is asking. The Redundancy Theory as expounded by Ramsey, amongst others, suggests that the use of the notion of Truth is unnecessary, since to say that p is true is simply to assert that p. Pinning down which question the Redundancy Theorists are answering is actually pretty hard, but seems to be something to do with what rôle the use of the concept 'Truth'plays in language - and they are suggesting that it only has a redundant one. Such a theory does not say that it is necessarily a total waste of breath to use the word since often it can add clarity and simplify a complex notion - such as saying that a valid argument preserves truth rather than that a valid argument, if it takes a premise which is, results in a conclusion which is too, or perhaps that if a valid argument takes premise p and p then if the conclusion is c then c. Overall, however, such a notion, they claim, does not provide any more information than p when saying 'p is true'.
The Redundancy Theory, by suggesting that statements involving Truth can always be rewritten without involving the concept, is able to answer a number of problems plaguing ideas about Truth. The Liar Paradox (where someone says 'This statement is false') is rendered academic. If we are to translate 'p is true'simply as 'not p'then what he appears to be saying is simply 'Not this statement', or 'It is not the case that this statement', which simply begs the question 'It is not the case that this statement is what?', and is not in fact a well-formed sentence. This theory does show something interesting - namely that the predicate 'is true'can only take a whole statement as its subject - so '2 plus 2 is 4 is true'or 'Edinburgh is north of London is true'both work whereas 'Blue is true'doesn't, and this ties in well with the theory since clearly 'Edinburgh is north of London'and '2 plus 2 is 4'stand on their own as sentences but 'Blue'does not. Likewise, 'This statement'does not stand on its own, and therefore can be neither true nor false. It should be clear, therefore, that a sentence needs to have a main verb already before it can be said 'to be true', which does not act as a first order predicate in its own right.
It might also be sensible to suggest that there is another use for 'is true'that the Redundancy Theorists seem to miss - that it can be used to assert that a particular sentence is an objective rather than subjective matter. We might, for example, say 'John's claim that it is hot today is neither true nor false', because the matter is subjective, and the Redundancy Theorist would have a harder time rephrasing this without the use of the words 'true'and 'false'('It is not the case that it is hot today or it is not hot today'?). Such analogous phrases (though some might argue that these cannot be used interchangeably) for 'is true'such as 'It is true that…'or 'It is a fact that…'are clearly often used rhetorically to assert that something is objective, rather than subjective - 'It is a fact that lots of people died on the Titanic'might be a response to someone that suggested it's just a matter of opinion. It may, however, simply be that such a sentence is able to assert two things concisely, and that 'John's claim that nobody died is true'is actually just saying 'John claimed that nobody died. Nobody died', and that it is a coincidental side-effect that if it is said that 'Nobody died'(presuming the voice is an omniscient narrator, rather than a person expressing an opinion) then this must be an objective fact.
The question answered, however, by Redundancy Theory is not the same question as is answered by most other theories of Truth, and does not usually compete with them rather than complementing. I don't think that there are any theories that would seriously suggest that it could be that 'if p then p'but not 'if p then p is true'. To assert that 'p is true'just means p might be the case, but that doesn't answer the question of what it is for it to be p, rather than for it not to be p, which is one of the questions that many other theories seek to address. It might be claimed that this question had been answered, that 'is true'means nothing, therefore things cannot be said to 'be true', yet this certainly hasn't been shown by the Redundancy Theorists argument. Clearly there is still some discussion to be made as to what is special about p that means that p rather than not-p.
The most simple (to understand, at least, not necessarily to explain the intricacies of) theory as to what it is for it to be the case that p rather than not-p is that of Correspondence Theory, which argues that for something to be true is simply to say that it corresponds to a fact in the world. It is not clear what the nature of such a correspondence would be, or indeed what type of thing the facts would be, yet this theory has a clear advantage in its simplicity. It is when pressed as to exactly what is meant by the theory, however, that Correspondence Theorists hit upon problems.
Haack notes the important difference between definitions and criteria for saying that something belongs to a particular set, and suggests that whilst a definition might be interesting - such as that measles is caused by a particular bacteria - attempting to work out what conforms to a definition is usually as hard as the task of deciding what belongs in the class it defines without any definition. With criteria, however, whilst we may not be able to say that something belongs to a class with as much certainty as when we say this is so by definition, we will have a much easier task labelling the majority of cases as conforming or not to simple criteria - such as exhibiting many red spots when you have measles. Criteria for truth, therefore, may in fact be preferable to an actual definition, since it can get you a good deal further, rather than leaving you with as hard a task as you started.
Russell, for example, argues that 'P consists in its correspondence to the fact that p', yet, as Haack points out, this is too close and obvious and thus doesn't seem to be saying anything much at all. Of course if p is true, then there is a fact that p, yet this seems to be just a matter of playing with synonyms rather than pointing anything out, and it is unclear what advantage there might be in noting the 'fact that p'on top of just p. This might be compared to Austin's view, which certainly is saying more than Russell's (as is far more controversial for it); he says that to say p is true is to say that p corresponds to the facts being as p says. This, a far more instrumentalist definition, which seems to involve only the demonstrable qualities of that which might be affected by p being true or not in discussions of the meaning of the claim that p, seems set to offer a much easier epistemological task, since only that which is observable affects truth value - and this means that such a view is favoured by the positivists.
The complaint levelled at Russell, however, seems to ultimately be the problem with Correspondence Theory - that ultimately it's all very well saying that p is p because it relates to a fact that p in the world, but this doesn't seem to be saying much about the nature of the facts themselves, and surely these are the things that make up the Truth we are trying to find the meaning of. Even Austin's view, which attempts to avoid such a problem by suggesting that any claim that p will relate to a host of facts, all of which would have to hold if p, simply sets the question a further stage back, and we still have to ask what is the nature of this set of facts, and how they relate to claims such as p.
Such problems led to the suggestion (particularly by Bradley and Rescher) of Coherence Theory, which, again, does not have to be a competing one, but can answer some of the practical questions that Correspondence Theory cannot, such as how we might come explain our own formulation of truths. Bradley, for example, suggests that whilst Correspondence might potentially be the definition of Truth (and as we have seen this could be so compellingly obvious a proposition that we might not wish to deny it) we should use something he calls coherence as the test for truth. Bradley (somewhat controversially) suggests that coherence is not just a matter of consistence but also of comprehensiveness - namely that we should prefer the least complicated, most consistent and (some would say) most elegant set of beliefs. I would suggest, however, that this suggestion is so plausible a description of how we do in fact act that it will probably not be terribly sensible to use it as a test; we are hardly likely to be attracted by a complicated, inconsistent and inelegant solution so unless we are going to say that Coherence Theory is successful because the world is essentially coherent all we seem to be doing is patting ourselves on the back for the great technique we currently use, with no basis to do so. It also seems to miss the point since things might well not be as we perceive and surely if there is any weight in Correspondence Theory it requires that things could correspond in a way we might not expect from the current evidence.
Rescher, however, avoids this problem by suggesting that all Coherence Theory attempts to do is to distinguish possible truth sets for those that simply won't work (due to lack of self-consistence) and doesn't try to pick one - escaping questions about how we judge which 'the most consistent'or 'least complicated'set of beliefs might be. This, however, is exactly what Russell objects to, claiming that Coherence Theory is unable to distinguish truth from consistent fairy tales - yet I'm not sure such a response would be fair, since Russell offers no better explanation of what correspondence with the truth consists of, and many Coherence Theorists do hold correspondence to be the actual defining factor, even if we currently lack a concrete conception of what correspondence consists of.
Another important formulation of the question 'What is Truth?'might be that it is asking what use Truth is / might be put to. The best answer to such a practical question, which ignores the most severe metaphysical questions, in favour of an explanation of what place Truth has in how we live our live is named, appropriately enough, Pragmatism Theory. This theory essentially says that Truth is utility - that we use our other knowledge and experience to find assumptions that work best, i.e. produce the most useful results . There is a very attractive aspect to this theory - namely, as Haack puts it, that 'truth is satisfactory to believe'; what she means by this is that we can trust a truth if we have one, and rely on it, simply because truth is defined as that which, if relied upon, will lead us to positive results. Unfortunately, such a conception of Truth seems implausible - whilst people may wish to rely on truths we do not usually hold that they may automatically do so merely by definition.
If a Correspondence Theory is correct, however, we would still be able to rely on truths (since, indeed, the facts would make up the relevant information about a situation and surely we have to believe that full knowledge of the facts is a better preparation than ignorance) so is not incompatible with Pragmatism Theory, but does not supply such a quick route to truths, and in fact makes no guarantees that we could ever know a truth on which to rely. If we do know of any truths, however, we must surely be able to rely on them, and it is in this simple concept that I think Pragmatism Theory does turn out to be one of the most useful answers to the question 'What is Truth?'. James, for example, holds with the Positivists, saying that true beliefs are verifiable - since they are those which in the long run are confirmed by experience.
Russell, Moore and others, however, object to this theory because they claim that James and the other Pragmatists are simply identifying the true with congenial belief - yet James is entirely correct, I feel, in his claim that when we have more than one entirely consistent set of beliefs (though I would probably deny that we do ever have a set of beliefs where all are consistent - frequently we have contradictions, which is what the cutting-edge of science is involved in unravelling) we do use simplicity and elegance to choose between them, and thus go for that set which seems most useful in terms of preparing for the future.
Overall, therefore, I would argue that there are few conflicting theories of Truth, and that each seeks to address a different aspect of the issues involved. I would agree entirely with the Redundancy Theory, since it seems clear that 'p is true'offer no more factual information than simply p, and usually just has an aesthetic place in sentences in which it appears; I would argue the coherence and pragmatism are two ways in which we choose which sets of beliefs to reject and which to keep, but that ultimately only one of them probably actually corresponds to how things 'really are'. The one thing I would reject, however, is the Unanalysable Quality Theory - simply because there seems quite a bit of scope to formulate more interesting ideas about Truth, and we need not yet retreat to such a position of ignorance and admit defeat.