Time as a concept has intrigued philosophers for, well, a long time. Is it intrinsic to the Universe, or is it a human construct that helps us to describe the world around us in terms of equations? To the question “what did God do before the beginning of time?”, St Augustine is said to have quipped that He was preparing hell for those who dared to ask such questions. But ask such questions physicists must, no matter how dire the ever-lasting consequences. So here we go...

The idea that time flows in one direction is intuitive: we get older, greyer, balder and, no matter how much we may wish it, we can never go back. Physicists define this forward direction in the language of the second law of thermodynamics: in a closed system entropy never decreases — the Universe moves forward in time towards disorder. The problem is that the fundamental equations that we use to describe our world are not sensitive to this direction; they work equally well with time going forwards or backwards. This is Loschmidt's paradox, named after the nineteenth-century Austrian physicist and chemist who said that irreversible processes should not emerge from time-symmetric dynamics.

Credit: © ISTOCKPHOTO/BELTERZ

Lorenzo Maccone now proposes a way to reconcile our everyday notion of time with quantum mechanics (Phys. Rev. Lett. (in the press); preprint at http://arxiv.org/abs/0802.0438v2, 2008). His basic idea is that changes that involve an increase or a decrease in entropy can both take place, but the decreasing cases do not leave any lasting trace: “the only physical evolutions we see in our past, and which can then be studied, are those where entropy has not decreased.”

The caveat for the second law of thermodynamics is that all systems must be uncorrelated. However, correlations between us, as an observer, and other systems do exist even if we are not aware of them. A process that leads to a decrease in correlation would lead to a reduction in entropy; however, the observer would not be aware of them, as memories are correlations and would have been erased. Maccone notes that even a super-observer who can follow all correlations will not see an increase in entropy. St Augustine would not be impressed!