On using the multiverse to avoid the paradoxes of time travel

Abstract

To avoid the unpalatable consequences of unidimensional time travel the celebrated paradoxes of preventing one's own birth, etc. it might seem reasonable to appeal instead to the theory of the multiverse.' On that theory, rather than having a time traveller, S, journey into the past of his own universe (call it '(') and thereby court the apparent contradictions, we can instead posit an alternative universe ('P') as his proper destination. The event of S's departing the year 2001 for the year 1001 (say) will therefore be more precisely described as S's leaving his own universe and time, 2001a, and travelling not to 1001a, the ancestor of his own time, but rather to 1001p, i.e., the year 1001 in another universe.2 It follows, say the advocates of the multiverse, that any action S undertakes after travelling to the past occurs in an alternative universe and so cannot affect and, a fortiori, contradict the events of his own. What is intriguing about time travel under the aegis of the multiverse is that it seems, at least prima facie, to allow for the reality of tense; it does not seem to require, in other words, that time be what McTaggart (1908) calls a 'B-series'. The traditional objection against accepting both the possibility of time travel and the reality of tense is well known: if past and future are unreal (as a genuinely tensory series, of course, requires) there will be no time, properly speaking, for the time traveller to travel to, and thus if tense is a genuine part of the world, time travel cannot be. This problem, however, does not arise on the theory of the multiverse as long as we hold the following principle: for every time t, there exists a universe, o, such that the present time in o is t.3 Only a principle such as this could