The first distinction: observable universe versus whole universe
The observable universe is the region from which light or other signals have had time to reach us. It is not automatically the same thing as the whole universe. A region beyond our current particle horizon may still belong to the same connected spacetime and obey the same large-scale cosmological solution. Calling every unseen region “another universe” therefore creates confusion before the physics even begins.
Where a cosmological multiverse can enter
In some models of eternal inflation, local regions stop inflating and reheat while other regions continue inflating. The resulting reheated domains can become causally separated. In that context, “pocket universe” or “bubble universe” refers to a region with its own post-inflationary cosmological history embedded in a larger inflating spacetime. This is a model-dependent construction, not a directly imaged population of external worlds.
Not every multiverse idea is the same
Inflationary pocket universes, a landscape of metastable vacua, Everett-style many-worlds quantum mechanics, and mathematical-universe proposals answer different questions and are not interchangeable. A careful discussion must state which mechanism is being considered and what observational consequences, if any, follow from it.
What is actually observed?
We observe one connected cosmic region: the universe accessible through our past light cone. We measure its expansion, the cosmic microwave background, primordial abundances, galaxies, large-scale structure, gravitational lensing and many other phenomena. No second universe has been directly observed, and there is no astronomical coordinate pointing to a confirmed external pocket universe.
Why scientists still study the idea
The multiverse becomes scientifically relevant when it is not inserted as a philosophical starting point but appears as a consequence of a specified dynamical theory. Then one can ask whether the underlying theory makes additional predictions, whether the relevant regime is internally consistent, and whether any observable relics are possible. The scientifically useful question is not “can I imagine many universes?” but “which equations produce them, and what else do those equations predict?”