[wahern]

The problem is that there's no generic way to describe the universe without contextualizing the characteristic you're interested in. But as a layman (not an astronomer, not a physicist, not a topologist), I'd argue that it's fair to say that the universe is infinite. I say that not only because of where the evidence regarding expansion and geometry of the universe points, but because even in a discussion of the visible universe you have to address the fact that the visible universe is shrinking as objects at the edge disappear due to expansion. Those objects don't cease to exist (at least not unless you make some highly contentious metaphysical arguments), which means there's no avoiding the inference that there can be (and likely are) an infinite number of objects which exist but which are not visible.

[TheCapeGreek]

I'm probably missing something, but my understanding of the "visible universe" is that the light from further out has yet to reach us, and this the visible portion should be constantly expanding (at the speed of light). Then, how would objects at the edge disappear unless the expansion is faster than light?

[cygx]

Objects don't outright disappear: As with other event horizons, they freeze in time and become redshifted as they approach the cosmological one (they also become fainter due to an increase in proper distance).

The cosmological event horizon is the light cone at future infinity and the asymptotic boundary of the observable universe: Light emitted within the horizon will take a finite time to reach us, whereas light emitted right at the horizon would take an infinte amount of time to arrive; in a way, light emitted beyond the horizon still moves towards us in the sense that the comoving distance decreases, but we'd have to wait a longer-than-infinte amount of time for it to arrive...

[maze-le]

Because they are beginning to move faster than c, when these objects are at the very edge, due to the space between us and the edge expanding faster than c. This is called the cosmic horizon of the universe - sort of like an inside out event horizon of a black hole. A particle inside the horizon with a speed of c can still reach us (albeit very strongly redshifted), a particle outside the horizon would have to travel faster than c to reach us. The thing is, our own universe is still not old enough, that it can have a proper event horizon. It is estimeted, according to a model that assumes dark energy is a cosmological constant, that the universe must be at least 16GY old for a cosmic horizon to develop.

Oh, and by the way, an expansion faster than the speed of light is consistent with relativity. Special relativity only describes local laws of physics -- you and the edge are not "local". And general relativity doesn't have a constraint on a maximum velocity between 2 arbitrary points in spacetime.

[akvadrako]

The expansion is FTL.

[ubershmekel]

Objects becoming invisible does not infer infinite objects. When you observe shopping carts leaving the store do you assume there are infinite shopping carts outside of the store?

[wahern]

True, but my point was that you already have to contend with the reality of objects that exist but which are beyond the visible universe. Where are all those shopping carts going? You can't hide the ball there without eliciting even more questions than you answer.

Limiting discussion to the visible universe (to a "finite" universe by eliding messy details) can mislead by creating seeming contradictions. It's sort of like saying that evolution doesn't exist (as a first order approximation) because the lineage from ape to man is just too complex and doesn't really matter; let's simplify things by eliding that lineage so we have an easier time analogizing human morphology and genetics as it relates to practical questions. It can work superficially but even laymen will have a sense that things don't add up, not to mention that it doesn't help resolve the more important "big" questions often implicit in any discussion.

The fact that the universe is likely infinite stems from experimental results confirming topological characteristics that reflect infinite space. Fortunately, when you try to conceptualize phenomena like the Big Bang, a flat, open infinite universe actually makes things simpler, IMO.