# Can you have a one-way street between two places in the same flat space?

Event horizons only allow passage in one direction. As far as I know, in black holes the event horizon separates two regions of space, and in some cases the inner region can even be asymptotically flat. That's an example of a one-way passage between two regions of flat space. What I am asking about is something one step further than that: can you have a one-way street between two asymptotically flat regions of space which are also connected by a flat bit? If that wording isn't clear, I'm imagining a situation where there are two paths from point A to point B, but only one path from point B to point A, where one of the A$$\rightarrow$$B paths goes through a one-way bit (not reversible), and the other goes through normal space (and so can be taken in both directions).

• You'll get finer answers from physicists, but my impression (a well-read layperson's) is that the passage of any "outside observer" across an event horizon takes an amount of time not distinguishable from an infinite one, which is why real (i.e., astrophysical) black holes were 1st called "frozen stars". Curvature characterizes the shape of any star, and it's their gravitational collapse (when the nuclear fuel whose conversion to energy is exhausted, ending the pressure which had prevented it) which creates the "hole" and starts the propagation of the E.H. outward from the star's center. May 4, 2020 at 0:38
• So an irreversible curvature seems essential to the natural prevention of perceptible passage even in the inward direction across an event horizon, at least for creatures like us, whose shape allows them to recognize curvature: It's possible that extremely flat creatures might contain intelligence which would prevent them from perceiving the curvature but nevertheless allow them to discern some of its effects, but their nature would be outside the purview of a physics site in "our" world. May 4, 2020 at 0:40
• There are many types of horizons: the event, apparent, trapping, isolated, dynamical, evolving, causal, Killing, non-Killing, universal, Rindler, particle, and putative horizons, etc. The event horizon by definition is a boundary beyond which events cannot affect an observer. So your question should be clarified as being about the apparent horizon, not about the event horizon. Perhaps you should look at the Rindler horizon to see if it is what you are looking for. May 4, 2020 at 6:44
• Since Wikipedia's treatment of the Rindler Horizon contains unusually lengthy formal notations, I've found a site at gregegan.net/SCIENCE/Rindler/RindlerHorizon.html that abounds in graphics and analogies: It's by a fiction writer named Greg Egan and does not qualify as a resource recommendation by PSE standards (leaving this comment subject to early deletion), but its author has co-authored one scientific paper by a well-known physicist (John C. Baez), so its analogies and graphics may provide a tenuous answer to your question. May 4, 2020 at 15:45