So traditionally, philosophers have often talked about ‘truth’ in something like the following way. A ‘thought’ is true when it corresponds to reality, and false when it does not.
Some people attack this ‘correspondence theory of truth’, but I’m not going to. It seems like quite a basic feature of what we mean by ‘true’ that true thoughts are ‘right’ or ‘accurate’ in the way they represent things. It’s not that which I want to take issue with.
Rather, note some of the structural features of the traditional account. There are two states, true and false, and they basically divide everything between them (even if they’re allowed to have intermediate states of ‘partially true’ or ‘close to the truth’), and they refer to individual thoughts.
How does this sound as an alternative: the opposite of, the complete absence of, truth (accurately corresponding to how things really are) is not ‘falsity’ but rather ‘fantasy’. If I’m thinking about a giraffe with a runny nose, and my thought ‘connects up with’ a real giraffe, and that giraffe does in fact have a runny nose, then my thought is true. But if there isn’t such a runny-nosed giraffe out there, my thought is just fantasy – I’m just imagining a bizarre animal.
So what’s the difference between fantasy and falsity? This is where the other strucural feature comes in – the individualism or atomism of the traditional view. I would suggest instead that thoughts must be looked at in relation to the ‘surrounding’ thoughts, the other things that a person thinks which are (in some vague unspecified sense) ‘close to’, that thought.
Falsity, I think, is an untrue thought surrounded by true thoughts. It is an untrue section of a larger thought-structure which is, by and large, true.
To go back to that giraffe: if I have a lot of very accurate thoughts about the giraffe in question (I know where it lives, I know what it’s called, I have some way of identifying it), then and only then can my thought that it has a runny nose be properly false: the true thoughts allow the nose-related thought to ‘get a fix on’ or ‘connect up with’ the actual giraffe, i.e. allow it to be ‘about’ that giraffe, so that we can then compare the way the thought represents that particular gireffe’s nose, and the real nose.
If it now turns out that the giraffe has in fact got a very healthy nose, then my thought that it has a runny nose is false. But if that same thought had been surrounded by other ‘false’ (i.e. not true) thoughts, then it would just be fantasy. Obviously this suggests that there can be degrees of fantasticalness, and conversely degrees of truth-and-falsity-ness, i.e. between a single thought totally surrounded by accurate ones (which is either true or false) and a thought totally surrounded by inaccurate ones (which is fantasy), there will be many intermediate degrees.
An example of how this would be applied to some problems in analytic philosophy of language:
Definite descriptions. Some philosophers (and only philosophers) have argued very emotionally about whether sentences like “the king of france is bald” are false or meaningless. It seems slightly wrong to say (as Bertrand Russell did) that they are simply false, in the same way that “the queen of England is bald” is false – because there is no hairy king of France. But it also seems too extreme to say that they’re meaningless, because the key phrase (“the king of france”) doesn’t refer to anything, since they’re so clearly meaningful and intelligible.
It seems to me that this dispute is a product of epistemic atomism: because the statement, and the thought it expresses, must be taken on its own, the fact that one of its ‘surrounding thoughts’, one of its ‘presuppositions’ (that France has a king), is inaccurate, cannot be properly dealt with – the slightly fantasy-like character of the thought expressed can only be described by the extreme option of branding it ‘meaningless’.
In place of this, the situation becomes easy when we recognise that thoughts can refer, and can be true or false, only against a background of other thoughts, and when these thoughts are faulty, the thought in question will not quite be either ‘true’ or ‘false’.
Similarly, I feel this sort of revision might be fruitfully applied to:
-the Gettier cases and the definition of knowledge;
-Russell’s concerns about proper names and ‘illusions of thoughts’;
-co-referential names and the paradoxes they present;
-the limitations of Kaplan’s tripartite analysis of meaning into character, content, and reference (and its ancestor, Frege’s division into sense and reference).
So in conclusion: espistemic holism yay! Thoughts depend for their connection to reality on their connections with other thoughts; a whole big group of related thoughts, none of which are true, is not falsity but fantasy; full-strength falsity only works against a background of truth.