The study of cosmology, the branch of the physical sciences that investigates the universe and its properties, presents quite a practical challenge: contrary to most other sciences, where different samples can be probed and analyzed directly, it's impossible to experiment with different universes in the lab.
We have our own example, this vast expanse of space and matter we live in, and that's it. So, to make sense of the cosmos, we study what's in it — the different types of matter and their properties such as temperature, density and pressure, and how it's distributed in space. To that we add observations on the size of the universe and how it evolved in time, and so on — and from this collected information try to come up with plausible explanations that describe what we see.
Some physicists, such as MIT's Edward Farhi, Alan Guth and Jemal Guven have even considered whether it would be possible to create baby universes in the laboratory. [The technical reference is Nuclear Physics B 339, 417 (1990).] The idea is to shrink a chunk of matter to incredibly high densities, forcing it to become a black hole. Occasionally, the authors speculated, this ball of matter could branch off to create a baby universe inside the black hole. Amazingly, this branch could grow up to a large size without interfering with the lab so that the cosmic Dr. Frankenstein wouldn't destroy himself and the rest of civilization. (Very roughly, think of it as a tunnel dug into the ground and away from observers at the surface.) Others, like the late cosmologist Edward Harrison, speculated that a super-advanced civilization created our own universe in a lab. This would explain why our Universe is conducive to life and so on.
But let's steer back to more palpable reality. There are serious questions concerning the viability of making baby universes in the lab, in spite of how cool the idea sounds.
Scientists use two ways to study the universe: we collect information directly, observing every object that we can such as galaxies, stars, clusters of galaxies, black holes, cosmic ray particles, etc. And we can simulate the universe in the lab: we may not be able to create a universe in the lab, but we can recreate parts of its history. There are two kinds of "labs" to do this: particle colliders such as the Large Hadron Collider (LHC) at CERN in Switzerland, where the Higgs boson was discovered in July; and in computer simulations that examine how chunks of matter and radiation behave under the laws of gravity (and possibly other forms of interactions such as electromagnetic if there are charged particles or magnetic fields present.) Today, let's stay with particle accelerators.
How to simulate the cosmic past in a particle accelerator? Recall that, according to the Big Bang model, in the early stages of its history our universe was very hot and dense: the matter that we see today in galaxies and stars was dissociated into its most basic constituents, electrons, quarks, and other elementary particles. (Quarks are the components of the familiar protons and neutrons.) From oranges to atoms to protons, every chunk of matter is held together by attractive forces; earlier on in the cosmic history the heat and density was such that these bonds broke down and particles were free to move about. Whenever quarks tried to get together to make a proton, collisions with other particles would interfere. As a result, protons formed at about one millionth of a second after the "bang," when the universe had cooled enough to allow quarks to get together, while the first hydrogen atoms (a proton and an electron) only formed at about 400,000 years after the "bang."
When scientists at the LHC make protons move almost at the speed of light and then collide them against other protons (or heavy atoms and other fixed "targets"), the energies of the collisions are so high that they reproduce, for fractions of a second, the conditions prevalent in the young cosmos. This way, physicists are essentially traveling back in time to study the cosmic infancy in a controlled fashion. The study of the very small informs the study of the very large.
Some recent results from collisions of protons with heavy atoms found very strange results. Identical particles that travel away from the collision point in opposite directions do so in mirror-like paths, as if keeping their connections even at large spatial separation: it is as if they "know" where their partner is going. This effect, quite possibly the same mysterious entanglement observed in quantum systems such as in pairs of photons and atoms, had never been observed in the much more violent particle collisions with heavy atoms.
Could this kind of behavior have played a role in the early universe? One of the greatest joys in scientific research is to be surprised by nature. And here we have a pretty big one, coming from the study of the smallest bits of matter, which may force us to rethink a thing or two about the physical properties of the cosmos.