Dummett, Michael - A Defense of McTaggart's Proof of the Unreality of Time

08/28/2009

Truth and Other Enigmas, Harvard University Press 1978

This article gives a lengthy explication of McTaggart's argument for there being a contradiction in time. Author reconstructs the argument:
a) an event M is past, present or future
b) an event M is before, at the same time as, or after another event N
Crucial here is that the facts of kind a) cannot be reducible to facts of kind b). Apparently McTaggart's reasoning for this is that change is essential to time, but there is no change in the truth function of the fact 'M comes after N', or 'M came before N'. Such facts, if true, are eternally true and therefore don't involve change. (pg351) Now the problem is that an event M can be described using three mutually incompatible predicates (past, present, future). This leads to a contradiction in describing the event, thus time is an unreality.

Author considers a possible response: the predicates applying to event M isn't just 'past', 'present', 'future', but 'will be past', 'is present', and 'was future'-- building tenses into the predicate. But adding tense verbs into the predicates only pushes the further back, since now there are 9 different predicate combinations: 'was past', 'is past', 'will be past' and so on for 'present' and 'future'. In any case the contradiction can be generated again. (pg351-2)

A more sophisticated response employs 'token-reflexive' expressions (e.g. 'I', 'here', 'now'), which are like indexicals that change a sentence's truth conditions according to context. We then specify that token-reflexive expressions are essential to time-facts and resolve the apparent incompatibility (pg353). This response, author points out, trivializes McTaggart's argument since it could also be made for space or personhood, just as it was for time. Yet McTaggart doesn't make analogous arguments about space or personhood, so author thinks there is more to his argument.

The difference, author claims, between the argument about time and an analogous one about space is that token-reflexivity is essential to time but it isn't to space. (pg354) Author discusses how it is possible to have a description of objects in space that doesn't include the describer-- that has no 'spatially token-reflexive expressions in giving a description of the physical uinverse' (pg354)-- but the same is not possible for time. Author argues that any account of a sequence of events without token-reflexive expressions will leave the question 'but what is happening right now' unanswered, therefore the account will be incomplete. Another possibility is to think of an external observer seeing time as a fourth dimension that is static, similar to how a full spatial description of the world would be a static image in 3 dimensions. In a weird turn, author argues that what would be observed would be a 'model' of events, not the events themselves. (pg355)

The possible response to this essential aspect of time is to jettison the notion that there is one complete description of reality (pg356), but author admits he is partial to this. Instead, it appears the unreality of time is self-defeating, since our apprehension of the world is certainly temporal.