Saint Augustine said that he knew exactly what time is, up until someone asked him to explain it. The question of why we remember the past, and not the future, is a question like that, and is in fact a subset of it, since of course it deals with questions about time. At first it might seem like one of the questions that is only of interest to those interested in substance consumption, along the lines of asking how we don't know that our entire universe isn't some big atom in some larger universe, but it actually has been addressed seriously at various points: for example, in Stephen Hawking's popular work A Brief History of Time.
One of the reasons that the question comes about is that the laws of physics should be invariant in both directions. Any physical process, if totally reversed, is still allowed. In a world of totally elastic, random particle motion, the forward motion of time would be meaningless. Yet, in four ways, time seems to have a direction:
- In certain rare particle interactions, that are only observed in very specific circumstances.
- In the overall thrust of cosmological time, the fact that the universe does change overall from the Big Bang until now.
- In the fact that almost everything we see increases in entropy, that eggs break and stars run out of fuel.
- And, most mysterious, in the fact that we have a psychological arrow of time, that we, as such, remember the past and not the future.
So the question is, how are these things related, and can we even begin to understand such a question? Dr. Hawking seems to think they are, and attempts to give an answer: we have a psychological direction of time, because it is tied in to the universe's overall direction of entropy. In other words, we remember things because when our brain increases in order by remembering some new information, untold amounts of sugars have to be disordered back into carbon dioxide and water. Our brain experiences the same disorder as the rest of the universe, and that is why it understands time as going in that direction.
It is not a very adequate answer, but hardly a bad one considering the scores of philosophical issues that have to be addressed, as well as the fact that Hawking is going to have some predisposition for reductionistic approaches. If we look at the brain as a collection of atomic units (atomic in the philosophical sense), that can have settings of either "on" or "off", which is a simplified version of how neurons behave. It is problems with this model itself that lead to problems determining why the mind should be set on either the future or the past. A neuron can be set into firing by a number of conditions, some of which might reflect something that has happened previously, upstream along the lines of neurons or in the sensory realm. However, the neuron itself doesn't know what its state "refers to". As far as the neuron "knows", it is either on or off. The fact that downstream neurons were on or off at earlier times, or that sensory data arrived earlier, can not be stored by a neuron that only can encode one piece of information. This is of course, only a further problem inherent in the atomic paradox: how an object that has no parts (such as a philosophical atom or theoretical neuron) can store information at all. Even without taking the atomic paradox to its extreme, we still have to wonder how the brain, with its component of atomic neurons, can present a multiplicitious consciousness at all. Although "complexity" is often cited in discussing both the brain and the mind, it begs the question of how, barring any type of non-local effect, we can say the brain is complex, since each neuron can only have access to being either "on" or "off".
So then, Dr. Hawking says in effect that time flows forward for us psychologically because a neuron "knows" that sugar has become carbon dioxide and water. Which, to me, seems somewhat ridiculous, but not any more than stating that a neuron's state can intrinsically mirror any other change in the world, either in the immediate past, present or future. The fact that a neuron is switched on because, say, a mosquito landed on a forearm means no more or less than saying a neuron is switched on because the earth coalesced out of space. Neither of these can be intrinsically represented.
I find myself, in trying to explain why one particular answer to this question seems unlikely, to be drifting off from the original topic. As to why we remember the past and not the future, I can only say it is the past because it is what we remember. Beyond that, I do not know.