The Asymmetry of Causation

In light of (Pl )-(P8), the so-called causal asymmetry between a cause and an effect, which is traditionally formulated as ‘it does not follow from “x caused y”, that “y caused x’”, can be given in the following terms: if x produced y, y could not have produced x, because that would have required y to produce x prior to its own production, which is impossible.

This is a good time to point out a significant difference between my view and the perspectival view advocated by Huw Price (1996), who also argues that the standard view is biased by agency. I argue that the relation between interacting entities should be considered to be symmetrical, but the relation between the successive stages of the process of production to be asymmetrical. Price argues that the relation between earlier and later stages of a process hold a symmetrical relation, i.e. he argues that the state which a system appears to have changed from, and the state which it appears to have changed to, when its parts interact, really hold a symmetrical relation. Or perhaps better, he thinks that such processes have no intrinsic direction. Price’s argument is based on the fact that the mathematical formulations of the laws of nature in physics are time invariant. That is, because it makes no difference to exchange the time variable t with its contrary -t, in the fundamental laws of physics, any process can be described as going either backwards or forwards in time, without violating those laws.

Perhaps accounts of causality that rely only on the aspects of lawful connections between types of events in light of the most general formulations of the laws of physics will have to accept Price’s conclusion, but then, I think, those accounts are incompatible with the conception of causality as involving production in accordance with the genetic principle. If the relation between two states of one and the same system is genuinely symmetrical, neither could have produced the other. Admittedly, the genetic principle imposes no special direction to certain types of processes—a system can evolve from a state of type A to a state of type B, or vice versa—but it excludes that the relation between two token stages of one and the same causal process is symmetric. One will have to have come into being out of the other, or neither has. They cannot both come into being out of the other, because then both would have to exist before the other to allow the production of the other. Or, in other words, if a system evolves from A to B and back to A, say, when a caesium-133 atom oscillates between two energy states, then we still have a linear succession of token states sA(, sB„ sA;, sB4..., of which each state produces the next.

Furthermore, it seems to me that the time symmetry of the general formulation of the laws of physics disappears when you apply them to concrete systems, and therefore doesn’t really imply that time is reversible or symmetric. In the general formulas, the symbols for quantities are really just placeholders for whatever real worldly quantity you want to insert when you describe the behaviour of an actual system, and they therefore don’t have a direction. Once you insert the quantities representing the actual state of the system, they acquire a direction (they are vector quantities). As far as I know, the direction of these quantities is not affected by the reversal of the sign of the time variable. Imagine that you describe the state of a system at a given time whose parts are moving outwards from a centre, and then reverse the sign of the time variable in the equation you have, and then predict how that system will develop towards the ‘past’. It will expand towards the ‘past’, because your reversal of the time variable does not reverse the direction of the state of motion of the parts of the system. This means that the actual process you are describing will not turn into its mirror reflection if you reverse the time variable.

In other words, the time symmetry of the laws of physics only means that one physical system can evolve from state A to B, that another system can evolve from state B to A, and that a system can evolve from state A to B, and then back to state A, but it will then really evolve through a linear order of successive states 1,2, 3... An epistemic consequence of this is that we cannot tell in which direction the world is evolving merely by looking at the internal structure of physical processes. However, it seems to me that it is still the case that no matter in which temporal direction the world is as a matter of fact going, there is no way to reverse that direction. If it is evolving in what we at present identify as ‘backwards’, that is the only way it can be evolving (and therefore really is the forward direction).

It is then another question of how plausible it is to think that we could be wrong about the direction in which the world is evolving. What do we have to assume about the function of our brain to allow it to falsely represent the world as developing in the opposite direction from how it is actually developing? We can easily imagine such a scenario on a purely conceptual level; Martin Amis describes it in the novel Time's Arrow (1992). But can we explain how it really happens on rhe basis of our current knowledge of physical reality? A naturalist should conclude that our current physics rules out time reversal.

< Prev   CONTENTS   Source   Next >