Blogger and physicist Lubos Motl, discusses a recent preprint by physicists Roman Buniy and Steve Hsu (another blogger), on quantum entanglement, in which they ask not "why are things entangled", but why aren't things so entangled, given their common origins at the Big Bang, that it is impossible to distinguish quantum subsystems.
Motl, never one to mince words, argues that their implicit assumptions about what entanglement involves are deeply wrong, and hence that their paper is rubbish.
A discussion of the merits of the arguments of both expositions after a close reading takes more time than I have today. But, I will offer a few thoughts.
1. These are all intelligent people with PhDs in physics, who are also both talented at explaining physics ideas to the educated layman public, talking about very fundamental bread and butter quantum physics concepts, and reaching contradictory conclusions about how one should apply them in ways that have meaningfully different consequences. This doesn't happen in all areas of science. The fact that this happens at all, suggests that the debate of "interpretations" of quantum physics, is not, as often claimed, just semantics that doesn't mean anything to someone who really understands the physics at a professional level. Those intepretations (not just the standard ones with names, but the differing sets of rules for conceptionalizing and applying quantum physics concepts including entanglement), are meaningful discussions about non-equivalent rules of physics.
2. Another key point that the debate illustrates is that there is more to an equation than what ends up getting written down on a blackboard or in an academic paper. Equations are stylized summary of a set of relationships that routinely suppresses all sorts of assumptions from view from units, to applicable context, to a lot of other subtle interpretive and definitional issues. But, out of context and without the interpretive gloss that they are embedded in and arise from, they don't tell a coherent story.
3. Entanglement is a huge idea. Apart from it, quantum physics is understandable and useful, because every particle with a given set of quantum numbers behaves independent of its history in precisely the same way. No particle is an individual. Every particles is some subtype of one of fifteen kinds of particles, the fermions come in only about four variations each (plus momentum and location), and the bosons also have limited dimensions of variation (photons are identical except for frequency, polarization and direction; there are eight basic kinds of gluons, there are three basic kinds of weak force bosons).
Entanglement contradicts that assumption. Particles with a common history of the right kind are not independent of each other. A particle with one history is not an interchangable part with a particle with a different history. Moreover, while usually entanglements from more or less random histories of isolated particles in the past with each other average out and disappear at a macrolevel, there are circumstances where they don't and render an assumption of statistical independence when evaluating the behavior of the particles later in time unsound, sometimes less obviously than other times.
4. Then again, without digging too far into the morass, one of the core issues that Motl identifies is the notion of entanglement as a subjective property that is observer dependent, rather than an objective and universal property. A particle may be entangled as to one observe who does not know a particular quantum property of the particle, but not to another one who has observed it. This is foundational and important.