The Indistinguishable Origin

Did the universe begin? General relativity allows cosmological models with past singularities — moments where the mathematics breaks down, often interpreted as a beginning of time. The standard FLRW spacetimes, built on dust and radiation, have such singularities. The singularity theorems of Penrose and Hawking proved that under reasonable physical conditions, singularities are unavoidable. This has been taken as strong evidence that the universe had a beginning.

Linford (arXiv:2603.04159) shows that the evidence cannot support this conclusion. Every past-singular FLRW spacetime — every standard model with a beginning — has an observationally indistinguishable counterpart that either lacks the singularity or fails to satisfy the conditions required for a genuine cosmic beginning. The two models are different in structure but identical in every possible observation.

The argument extends the Malament-Manchak theorems, which establish that the global topology of spacetime is underdetermined by local observations. The extension is specific: it isn't just that we can't tell the full shape of spacetime from our vantage point. It is that we can't even tell whether the spacetime we inhabit has a beginning. The question "did everything start?" is observationally unanswerable — not because our instruments are too weak, but because the structure of general relativity permits twin spacetimes that diverge on this question while agreeing on every measurement.

The paper proposes two necessary conditions for a genuine cosmic beginning and shows that observers cannot gather sufficient data to determine whether either condition is met. The limitation is not epistemic (we could know but don't) but structural (the theory itself permits indistinguishable alternatives).

The through-claim: some questions about the universe are not empirical. Not because they are metaphysical — the question "did the universe begin?" has a definite answer in any specific spacetime model — but because the mapping from theory to observation is many-to-one. Multiple incompatible answers map to the same data. The beginning of everything is, from the inside, indistinguishable from its absence.