Monday, March 12, 2012

Can truth be defined?

To: Louis G.
From: Geoffrey Klempner
Subject: Can truth be defined?
Date: 23 April 2007 10:40

Dear Louis,

Thank you for email of 16 April, with your third essay for Possible World Machine, in response to the question, 'Can truth be defined? If you think that it can, give a definition and explain its philosophical significance. If you think that it cannot, what conclusions should the philosopher draw from that?'

As I read your essay, I was formulating in my mind the response that you eventually give: that scepticism and doubts about the possibility of overcoming clashes of perspective implies the existence of an objective truth.

That's not the last word, however, because the debate between the realist and anti-realist is left wide open. If we interpret the clash between theories or perspectives or ideologies in a sceptical framework, then we are assuming a 'truth out there'. However, for the anti-realist that is precisely the point at issue. Why should we see it in this way? Why believe that there is an ultimate scientific truth behind the competing theories, or an ultimate literary or historical truth behind the competing literary or historical interpretations, or an ultimate political truth behind the competing ideologies?

Let's look at your definition of truth which on the face of it is admirably simple. The truth is that which is.

The most obvious question is whether you are really prepared to embrace the view that the truth changes with time. As I write these words, the sky is very gloomy but it is not raining. A while ago, I could hear the rain. So my statement, 'It is raining' is false but a while ago was true. Is that an example of truth changing with time? No, because we are dealing with the familiar phenomenon of 'indexical reference', i.e. a statement with components whose meaning depends on the time of utterance. If I had said, 'It is raining in Sheffield at 10.30 am on 23 April 2007' that statement, if true, does not become false when the rain stops. In 100 years time, it will still be true that 'It is raining in Sheffield at 10.30 am on 23 April 2007.'

However, the anti-realist sees a serious problem looming here. Consider my statement, 'It is raining in Sheffield at 10.30 am on 23 April 1707.' It is possible that someone around at the time recorded this fact. But let's assume that the fact went unrecorded (Make the statement, 'No-one recorded the fact that it was raining in Sheffield at 10.30 am on 23 April 1707.') Don't we still want to say that the statement has a *truth value* even though because of the absence of evidence there is no way, in principle, to discover that truth value?

It is possible that your idea was that we should limit our ambitions and only define truth for the present, or at least the time round about now, thus avoiding the problem of truth 'changing' with time. However, this would still leave us with competing claims to truth, e.g. with respect to the political situation in Iraq.

Insofar as one wants to agree with your simple definition, this is because it basically says the same as Aristotle's definition: The statement P (for any P) is logically equivalent to the statement 'P' is true. The statement not-P (for any P) is logically equivalent to the statement 'P' is false. The word 'true' is just a way of removing quotation marks from a statement. If you gave a long speech and I agreed with everything you said, I could say, 'Everything LG said is true' and save myself the trouble of repeating everything. Or I could say, 'Nothing that LG said is true' or 'LG said something true'.

The view that this exhausts all that philosopher say about truth is known as the 'redundancy' or 'minimalist' theory of truth. (If you look these terms up on the internet you will find that these are not quite equivalent but the differences are not important for present purposes.)

The idea behind the redundancy or minimalist theories is that any attempt to say more (for example, to say something that would enable us to decide between realist and anti-realist views of truth) merely changes the subject. There is no *informative* definition of truth, nothing that will help us in our search for truths, or the struggle to decide which of two competing views is true. Some philosophers would regard this conclusion as, in effect, the claim that truth cannot be defined because it cannot be informatively or usefully defined. Truth is ultimately indefinable, because a formula which merely helps you identify what word we use for 'true' is not the same as a real definition, which gives the essence of truth.

An alternative view, is that the ''P' is true equals P' definition, when seen in its appropriate context is all we could ever want of a definition of truth. Armed with our definition, we can go on to tackle the semantic, epistemological and metaphysical problems which our minimalist definition of truth leaves unresolved.

This was a good, clear essay which raised a number of interesting issues.

All the best,

Geoffrey