It's an analogy, and as such it has limited applicability. Noone is "holding" space, it's infinite. The blanket is two-dimensional and embedded in the three dimensions of our world, hence it has boundaries. Space is, well, space. It's not embedded in anything. This is quite nonintuitive and here's where the blanket analogy hurts more than it helps.
The best way I know how to think about curved spacetime is by analogy to the Earth. We are effectively two-dimensional on the Earth's surface, as the radius of curvature is much larger than we are. Yet you can imagine drawing straight lines (i.e., great circles) on the ground to make (large) triangles and discovering that the angles don't add to 180 degrees and hence deduce that the surface is curved.
It's sort of the same way with space. Since we are three-dimensional, we can't directly perceive the curvature as we can an embedded two-dimensional object. A straight line would always look straight to us, yet we could measure the curvature by observing that the fundamental theorems of Euclidean geometry don't hold.
The Earth does slowly spiral towards the Sun. There's no "friction" per se, but it loses angular momentum to other stuff, like gravitational waves and the dark matter hslo. It just operates extremely slowly.
That depends on what you mean by "know", I guess... According to the current best estimates of the cosmological parameters, yes, it is infinite. The basic theory could be wrong, of course, and outside the observable Universe the properties could be different.
That would have been the case if the universe was closed, i.e. if the matter density was above the critical density. However, the matter density is only about 20% of the critical, which means it's infinite. (The remaining 80% of the critical density seem to be the cosmological constant, which means the universe has flat geometry to within experimental error.)
Comments
It's an analogy, and as such it has limited applicability. Noone is "holding" space, it's infinite. The blanket is two-dimensional and embedded in the three dimensions of our world, hence it has boundaries. Space is, well, space. It's not embedded in anything. This is quite nonintuitive and here's where the blanket analogy hurts more than it helps.
The best way I know how to think about curved spacetime is by analogy to the Earth. We are effectively two-dimensional on the Earth's surface, as the radius of curvature is much larger than we are. Yet you can imagine drawing straight lines (i.e., great circles) on the ground to make (large) triangles and discovering that the angles don't add to 180 degrees and hence deduce that the surface is curved.
It's sort of the same way with space. Since we are three-dimensional, we can't directly perceive the curvature as we can an embedded two-dimensional object. A straight line would always look straight to us, yet we could measure the curvature by observing that the fundamental theorems of Euclidean geometry don't hold.
The Earth does slowly spiral towards the Sun. There's no "friction" per se, but it loses angular momentum to other stuff, like gravitational waves and the dark matter hslo. It just operates extremely slowly.
Do we know that space is infinite?
That depends on what you mean by "know", I guess... According to the current best estimates of the cosmological parameters, yes, it is infinite. The basic theory could be wrong, of course, and outside the observable Universe the properties could be different.
I thought it was finite but unbounded.
That would have been the case if the universe was closed, i.e. if the matter density was above the critical density. However, the matter density is only about 20% of the critical, which means it's infinite. (The remaining 80% of the critical density seem to be the cosmological constant, which means the universe has flat geometry to within experimental error.)