To: Mark S.
From: Geoffrey Klempner
Subject: Reducing truths about the past to present tense facts
Date: 5th February 2010 13:00

Dear Mark,

Thank you for your email of 29 January, with your essay for the University of London BA Metaphysics module, in response to the question, 'All statements about the past are true or false in virtue of our present evidence of how things were in the past.' Discuss.

You have made a heroic effort to cover a wide span of views relating in one way or another to the question of truths about the past. All the various nuanced positions, however, in your view rendered unnecessary by the simple expedient of rejecting A-theory and being a B-theorist about time.

Looking at this question, I have difficulties even making sense of what is being asserted. You discuss the notion of 'presentism' and go off on various tangents including the idea of a Laplacian Supermind who is able to take in all 'present' (in the Newtonian sense) facts and deduce past and future on the basis of deterministic laws of nature. But this has got nothing to do with evidence. The Laplacian Supermind doesn't need evidence. It 'sees' the facts as they are now, directly. (Hence, questions arising from chaos theory are irrelevant because in a deterministic universe the only problem is the fact that all observations are necessarily approximate, plus or minus a given value, and the slightest degree of inaccuracy gets magnified a la Lorenz.)

You object to RAT on the grounds that, 'The major part of the past would simply vanish. For example, the dinosaurs for whom we have fossils would be the only ones that existed.' What on earth does that mean? How much of the 'truth' about a given dinosaur can you deduce from its bones?

The problem is that 'present evidence' is just that, evidence. When we 'deduce from evidence' what we are doing is induction, or abduction (inference to the best explanation). It is always possible that Martians manufactured and planted those bones.

This makes reductive 'anti-realism' a very different kind of animal. (I'm using quotes because I prefer Dummett's usage according to which any theory which delivers truth conditions is 'realist'.) When the RAT theorist says 'dinosaurs existed 500 million years ago', the meaning (=truth conditions) of that statement has nothing to do with flesh and blood creatures pictured in school books. We are talking about what we can find in the world around us, bones and stuff, and the various scientific procedures for determining their age such as carbon dating. That's what we mean. The pictures we may have in our heads have nothing to do with what we mean.

On this view, one doesn't get too fussed about how one defines 'the present'. 'Present evidence' is evidence which is more or less to hand, or 'present', in the sense that it is here, in a museum somewhere or dig-uppable. (Obviously, one can raise problems here. If the stuff is too far underground, then maybe it isn't 'present' -- however, there are different views possible here, for example, one can talk about what beings with superior technology could uncover etc. etc.)

You don't need to be a metaphysical 'presentist' in order to (be motivated to) hold RAT. I'm not rejecting your hypothesis that RAT is based on Presentism + Truthmaker Principle + Historical Truths Principle. That would be one kind, a rather obtuse kind, of RAT theorist. However, it seems to me much more plausible that RAT is motivated simply by verificationist scruples.

On the non-RAT view, statements about dinosaurs have a truth value which can never be conclusively verified. It's beyond our grasp. According to RAT, when we make statements about dinosaurs, either we are aiming our thoughts at truth -- at existing facts -- or we are just telling stories. Truth can only belong to verifiable statements. And what cannot have truth in this sense has no semantic meaning (even if it has 'meaning in our heads' or 'picture meaning'). If we are aiming our thoughts at existing facts, then these facts better had be accessible to investigation.

This view is one of the upshots of a once-powerful tradition in philosophy (logical positivism and the verification principle). In other words, this isn't so much about the metaphysics of time as the nature of truth and meaning.

Dummett's famous 1969 paper takes the argument to a new level. The motivating intuition is the same -- that semantic meaning isn't the same as 'picture meaning'. However, for Dummett grasp of semantic meaning is necessarily exhibited in rule-following behaviour, such as knowledge of verification or falsification procedures (the 'manifestation argument'), and in order to represent this knowledge we have to abandon the idea of a truth conditions theory of meaning in favour, e.g., of a 'verification conditions' theory.

Dummett would therefore reject the statement, 'All statements about the past are true or false in virtue of our present evidence of how things were in the past.' The statement implies the law of excluded middle: every proposition about the past has a truth value, and that value is determined by the existence of present evidence. When we say 'dinosaurs' what we really mean is 'dinosaur bones' (as it were). In any case, most historical statements are false. Whereas on Dummett's view we cannot meaningfully say, of any given statement about the past, that it is 'either true or false'.

On a historical note, discussion of these issues took up a considerable portion of my D.Phil thesis. What has survived of those historically distant investigations can be found in my book (ch.16), and also in the Pathways Metaphysics program.

All the best,

Geoffrey