I read the whole paper, but I still don't get exactly why this relation exists. They described some properties of "memory systems", and then said that a memory system must "remember more than one thing", and for some reason knowing the future as well as the past corresponds to "only knowing one thing", which means that anything that knows the future isn't a memory, because it only knows one thing.

Can someone please explain the whole thing with "knowing multiple things" and robustness to perturbations in the system?

10 upvotes because it has an interesting title. Anyone that attempted to actually read it could see theres no damn paper. Sorry I seem frustrated with this- I really wanted to read it. Time for bed.

Looks to be a rigorous exposition of an argument I saw to resolve Loschmidt's paradox (why there can be an arrow I time in a time symmetric universe). The argument is that formation of memories (whether in a brain or tape recorder or whatever) is necessarily an increasing entropy process, so all memories will be of lower entropy states.

IOW, it's not that "entropy increases futureward, we remember pastward, what a coincidence!" Rather, "if this were not an increasing entropy direction, we could not have memories of it."

First saw the point in Drescher's Good and Real.

Full-text is behind a paywall, but there's a preprint version available here that looks roughly the same: http://arxiv.org/abs/1310.1095

Thanks. We changed the url to that from http://journals.aps.org/pre/abstract/10.1103/PhysRevE.89.052....

There's a PDF button, which leads to http://journals.aps.org/pre/pdf/10.1103/PhysRevE.89.052102

EDIT: Actually, it appears that my university access may be providing the link. However, cing's link, though slightly different presumably due to being pre-journal formatting, is the same paper.

PDF: http://arxiv.org/pdf/1310.1095v1.pdf

Todd Brun is notable for being a collaborator (student?) of Jim Hartle for a number years when they worked on the decoherent histories formalism of quantum mechanics. Jim was one of the original developers of the formalism.