My feeling is that no a closed universe can't become open. Under normal evolution according to the Friedman equation the curvature would be more exagerated as time goes on (i.e. a a closed universe becomes more closed). Under say inflation the curvature can be flattened but i hesitate to say that it can't be flipped (i.e. go from closed to open), is this correct?


Topologically speaking, open universes and closed universes have a different topology ($\mathbb{R}^3$ versus $S^3$). Topology change would require the introduction of something bad (either closed timelike curve, loss of time orientation or naked singularities), none of which are compatible with FRW spacetimes (in particular, homogeneous spacetimes must be geodesically complete). Unless you're willing to drop the homogeneity and isotropy of space, one cannot change into the other in a continuous manner.

Edit : but wait! Consider a FRW spacetime, initially open, with a radius diverging in finite time. This would imply that the spacetime is indeed not causally closed, which may still allow for topology change. On the other hand, I'm not sure this makes sense as a spacetime, as if we consider two observers at the two poles of $S^3$, they would become separated by an infinite distance in finite time (and this is true of all points by homogeneity). It's difficult to consider as I can't find any coordinates where this scenario would play out without problems.

  • $\begingroup$ This doesn't sound like the theorems I've seen on topology change. Do you have a reference for this particular theorem? The theorems I've seen, by Geroch, Tipler, and Borde, say something more like the following. If topology change occurs and causal compactness holds, then causality violation occurs and the weak energy condition is violated. $\endgroup$ – Ben Crowell May 18 '18 at 21:25
  • $\begingroup$ True, I suppose it could also be a violation of causal compactness without naked singularities, though I can't think of such an example here that would be a homogeneous space at all times $\endgroup$ – Slereah May 18 '18 at 21:27
  • $\begingroup$ Perhaps some variant on AdS spacetime might work? $\endgroup$ – Slereah May 18 '18 at 21:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.