Fixed past, unfixed future

D. H. Mellor

pp. 166-186

in: Barry M. Taylor (ed), Michael Dummett, Berlin, Springer, 1987

Abstract

By "Fixed past, unfixed future" I mean that alternative futures are really possible in a serious sense in which alternative pasts are not. This proposition is of course a commonplace. It is not seriously contested and needs no argument from me. My problem is not to defend it, but to make sense of it on my tenseless view of time (Real Time, 1981). This is a problem because I deny the non-relational difference between past and future on which this difference in "fixity" (Mackie 1974, p.180) appears to depend; and the relational difference I do admit, between being earlier than some time and being later than it, hardly suffices. On Newton's deterministic theory of gravitation, for instance, a planet's position at a time is just as much fixed, i.e., determined, by its later positions as by its earlier ones: there is no asymmetry here between earlier and later to make tenseless sense of the past being fixed and the future not.