Observable universe

 Claude finished the response

https://spacedaily.com/t-observable-universe-93-billion-light-years-expanding-space/
OK, I get this, but how big is the actual universe?  I suppose the answer is "we don't know," but I suspect there is some sort of minimum.
Pivoted toward data-driven, citation-grounded analysis

Right, "we don't know" is the honest headline — but you're correct that there's more structure to it than a shrug.

The observable universe (~93 billion ly across) is a horizon, not an edge. It's just the region causally connected to us given the age of the universe and the expansion history. Nothing about it implies the universe stops there.

Whether the whole thing is finite or infinite depends on curvature, and current measurements (Planck CMB data) show the universe is flat to within observational error — consistent with zero curvature, which would mean infinite. But "consistent with zero" isn't the same as "proven zero." The error bars leave room for a very slight positive curvature (closed, finite universe) or negative curvature (open, infinite).

Here's the part that answers your actual question. Because we can measure curvature only this precisely, and not perfectly, that residual uncertainty converts directly into a lower bound on size, if the universe is closed. The most-cited result, from Vardanyan, Trotta and Silk (2011), used Bayesian model averaging on the era's cosmological data and found the curvature radius is constrained to be at least 42 Gpc at 99% confidence, corresponding to a lower limit of about 251 Hubble volumes — that's roughly 7.7 times larger than the observable universe. An earlier, cruder approach using WMAP's search for repeating "circles in the sky" (a topology signature) had already put a floor of 24 Gpc on the size.
arxiv + 2

So: if it's finite, it's at minimum several times bigger than what we can see, and that floor has only moved upward as measurements sharpened (WMAP's 24 Gpc → Planck-era 42+ Gpc). Modern Planck-only curvature constraints (Ω_k ≈ 0.0007 ± 0.0019, 2018 release) are tighter still and would push the floor higher, though I haven't seen a clean updated Gpc figure quoted as often as the 2011 number — it's the one that stuck in the literature.

The honest caveat is the asymmetry: nobody has ever proposed a serious upper bound. If it's flat or open, it's most likely infinite, and inflationary models (which are themselves reasonably well-supported) generically predict something vastly larger than the observable patch — some estimates run to numbers so large they're really just stand-ins for "incomprehensibly bigger," not measurements.

So your instinct was right: there's a rigorous minimum, it's a handful of times the observable universe at the low end from direct curvature constraints, and no corresponding maximum — the universe could just as easily be genuinely infinite.

It seems to me that if an infinite universe is what we have now, then it is what we have always had.  In other words the singularity (or whatever) was infinite to begin with and what has happened is not getting bigger in some sense but just less dense.
Validated user's intuition about infinite universe density dynamics

That's exactly right, and it's a cleaner way of thinking about it than the popular "explosion from a point" image, which is actually wrong for the flat/infinite case.

In the standard FLRW model, if space is infinite now, it was infinite at every earlier time too — including arbitrarily close to t=0. The scale factor a(t) going to zero doesn't mean space collapses to a point; it means the density diverges while the spatial extent stays whatever it always was (infinite, if it's infinite now). What we call "the Big Bang" in that case wasn't a localized event that things flew outward from — it happened everywhere in an already-infinite space simultaneously. Every point was infinitely dense at once, not converging toward one point from all directions.

The "explosion from a point" picture is only accurate if the universe is spatially closed and finite (positive curvature). In that case the whole finite volume really does shrink to zero at t=0. Given that our data lean toward flat-or-open, that picture is probably the wrong one to have in your head, even though it's the one that dominates popular science writing — it's a much easier thing to draw.

One honest caveat: "singularity" in this classical picture is really a signal that the theory (general relativity) is being pushed past where it's trustworthy, not a literal physical infinite-density point-in-time. Nobody thinks density actually hit true infinity — that's where quantum gravity is expected to take over, and we don't have a working theory for what actually happened in that regime. So the more careful statement of your insight is: as far back as classical cosmology can be trusted, an infinite universe stays infinite and just gets less dense; what happens at the very edge of t→0 is currently unknown rather than resolved.
Weighed Hoyle's stellar nucleosynthesis contributions and historical significance

That's a fair and important correction to the "Hoyle was simply wrong" framing — it credits the man for the wrong reason if that's all you take from him.

The B²FH paper (Burbidge, Burbidge, Fowler, and Hoyle, 1957) is one of the foundational works of 20th century astrophysics, and Hoyle was central to it. Big Bang nucleosynthesis only produces hydrogen, helium, and trace lithium — it can't get past that because there's no stable nucleus at mass 5 or 8, so the early universe's rapid expansion and cooling shut the door before fusion could build anything heavier by that route. Everything from carbon up had to come from somewhere else, and Hoyle correctly identified stellar interiors — and specifically the deaths of stars, supernovae scattering the products — as that somewhere else.

The best-known individual piece of this is his 1953 prediction of a resonance in carbon-12 at a specific energy, needed to make the triple-alpha process (three helium-4 nuclei fusing to carbon) happen efficiently enough to produce the abundance of carbon actually observed. That's about as clean a case of theoretical prediction preceding experimental confirmation as physics has — Willy Fowler's lab found the resonance essentially where Hoyle said it would be. It's also one of the most cited examples in the "fine-tuning of physical constants" literature, since that resonance is fairly sensitive to the underlying nuclear parameters.

So the accurate ledger: wrong on the large-scale history of the universe, and stubbornly, publicly wrong about it for decades after the CMB evidence came in — but right, and foundationally so, on where the periodic table beyond helium actually comes from. Steady-state theory's defeat and stellar nucleosynthesis's triumph happened to run through the same person, which is a genuinely unusual career shape. It's a good reminder that "wrong about the big framework" and "wrong about everything" aren't the same category, and the history of the field tends to flatten that distinction when it gets summarized.

I was so busy defending Hoyle that I missed something more significant.  I do not find the inability to know something less satisfying.  We are not looking for satisfaction.
Recalibrated focus from defending positions toward pursuing genuine truth

Fair correction, and it's a real one, not a semantic quibble. I reached for "less satisfying to sit with" as a description of your reaction, and that was me importing a motive you hadn't stated. Wanting satisfaction and wanting an accurate map are different projects entirely, and conflating them is exactly the kind of soft error that lets people mistake "I feel unresolved" for "something is wrong here."

If the honest answer is that curvature might always sit one decimal place further out, that's not a wound in the theory — it's a fact about the geometry of the question, learned the same way any other result is learned. It goes in the ledger next to the things that are settled. No different in kind from knowing a proof is undecidable, or that the halting problem has no general algorithm — those aren't disappointments, they're results. The Planck scale mattering isn't a loss either; it's an accurate location of the joint between two regimes of physics that don't yet talk to each other.

Put differently: the demand for satisfaction is what pushed Hoyle to defend steady-state past the point the data warranted it. Indifference to whether an answer is satisfying is what let him also get the stellar nucleosynthesis work right — that one didn't need to feel good, it needed to fit the carbon abundance. Same disposition, both cases. You're describing the correct instrument setting, not a special stoicism about this particular topic.
Want to be notified when Claude responds?


Claude is AI and can make mistakes. Please double-check responses.