Louis Ward

Fine-tuning is often taken to support the multiverse over a single-universe hypothesis, but this claim faces the ‘this universe’ objection. The objection argues that our evidence must be indexically specific, so the relevant evidence is that this universe is finely tuned rather than some universe. Positing many universes then does not make it any more likely that this universe is finely tuned. Requiring indexical specificity of evidence motivates a cosmological argument for the multiverse, since this universe’s existence is more likely on a multiverse than on a single-universe hypothesis. Goff ( 2024 ) replies that, because there are infinitely many possible universes, it is equally likely that this universe exists if there is one universe or many. In this paper, I present a trilemma against the single-universe hypothesis. First, if we accept both the requirement of indexically specific evidence and the assumption of infinitely many possible universes, then we have probabilistic confirmation of a multiverse. Any hypothesis with a finite number of universes predicts ‘this universe exists’ with a likelihood of 0, so multiplying this likelihood by its prior results in a posterior of 0. An unlimited multiverse predicts this evidence with probability 1. Provided this infinite multiverse hypothesis has a finite positive prior probability, an infinite multiverse is confirmed over finitely many universes. Second, if the indexical evidence requirement is rejected, the ‘this universe’ objection fails, and fine-tuning supports a multiverse. Third, if an infinity of possible universes is rejected, Goff’s response fails, and the cosmological argument succeeds.