Notes on Mellor, "The Unreality of Tense"
(from Real Time, reprinted in LePoidevin and MacBeath)

1. There seems to be a big difference between change through time and change across space, but for the tenseless view these are on a par. Related to this: the tenseless view will define change as having incompatible properties at different times. But this seems circular, since time in turn can only be understood as "the dimension of change." (That is, this is the only way to distinguish time from space, on the tenseless view.)

2. The tenseless view "reduces change to changeless facts." (? Well of course the facts don't change, but some facts are facts about the changes in things.)

On both these issues the tensed view at first seems to do better. (1) Difference between time and space: change is "the changing tense of things moving from future to past." There isn't a spatial analog. (We can construct sentences that are true in some places but not others, but these are completely reducible to space-less sentences: "it is raining here" is  true in some places but not in others, but "here" is not a real property that somehow moves from place to place.) (2) So change is not reduced to changeless facts; the earlier of two events is the one that was present first.

One might respond by suggesting that time and space are after all analogous; that change across space is just as real as change over time.

But the killer objection to the tensed view, for Mellor, is McTaggart's argument that it is self-contradictory. Trying to remove the contradiction by appealing to compound tenses does not help: these also contradict one another.

Mellor links the rejection of the tensed view with his token-reflexive analysis of 'now'. This holds roughly that "It is now the case that P" means "P is true at the time of this very token of the word 'now'."

Problems in embedded contexts.