So, at least as reported in the media, the physics community is still struggling with the problem of resolving the impossibility of retrieving information from beyond the event horizon of a black hole with the fact that QM seems to guarantee information is preserved.

However, as I understand, a distant observer will never actually see the creation of the black hole occur in finite time. In other words, over all finite times for a distant observer the situation can be described from via solutions to GR that don't actually result in an event horizon in finite time. Now if we assume that either the universe ends after finite time for that distant observer or the (psuedo) black hole evaporates after such time then all events can be described from some perspective without introducing any event horizon.

But by the equivalence principle in GR the perspective of any observer is just as valid and if information is preserved for any observer it is preserved for all observers. So how can there be any confusion?

(The assumption that the universe ends or the (psuedo) black hole evaporates in finite time ensures that there is no observer for whom the external observer's perspective is only valid for a finite amount of time.)

EDIT: Yes, as pointed out the equivalence principle in GR refers to something else. I mean whatever is the GR analog of the SR principle that tells us that a description in any inertial reference frame is equally valid. My understanding was that this was extended in GR to non-inertial observers but regardless of what it is called the point is that if there is any solution of GR that describes all space for all time without the introduction of event horizons then since this description creates no paradox there shouldn't be a paradox with respect to any description.

Also perhaps some assumption about the observer not accelerating to infinite velocity is required as well but maybe such observers are already not considered.

EDIT2: To be more precisce I want to choose a frame whose vector field matches the (usual) local coordinates for some observer who neither falls into a black hole or accelerats to infinite velocity relative to matter in the universe.

Perhaps I am confused and instead want to pick a maximal coordinate chart including all the points on the manifold occupied by my observer.

  • 3
    $\begingroup$ Please, please, answer only if you can provide some mathematical reasoning to support your argument. Resist the temptation to answer with what seems to you to be intuitively obvious - GR is rarely obvious or intuitive. $\endgroup$ Commented Jan 28, 2014 at 19:01
  • $\begingroup$ The equivalence principle in GR states that one cannot distinguish between a constant gravity field and constant acceleration. $\endgroup$
    – Jim
    Commented Jan 28, 2014 at 19:02
  • $\begingroup$ As soon as you talk about comparing things on the black hole horizon to the observations of distant observer, you are no longer talking about a local reference frame. $\endgroup$ Commented Jan 28, 2014 at 22:35
  • $\begingroup$ Sure, but I should still be able to define a (frame)[en.wikipedia.org/wiki/Frame_fields_in_general_relativity] which yields the correct local coordinates. The only importance of the observer far from the black hole who isn't falling in is to ensure that the frame doesn't end up crunching all time for such an observer into a finite period. $\endgroup$ Commented Jan 29, 2014 at 1:16
  • 1
    $\begingroup$ @PeterGerdes: No, you absolutely cannot do that. As soon as you have large enough charts to detect appreciable amoutns of matter, you will be able to detect that there is curvature, and the magic of equivalence is broken. The equivalence principle holds only so long as the manifold can be approximated by its tangent surface at a point. $\endgroup$ Commented Jan 29, 2014 at 2:56

1 Answer 1


This is answered very nicely here where Ben Crowel nicely explains the error I'm making (there is no global definition of time relative to a given observer that makes sense the way it does in SR).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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