"Eliezer S. Yudkowsky" wrote:
> If going
> into the future doesn't send me into a different timeline, neither will
> going into the past. There is no "temporal ether", no single direction
> of time, and the rules have to be the same in both directions.
> Causality *is* circular.
Hmmm, but there *is* an arrow of time, practically speaking, one that influences most any machine or process of significant complexity. The most obvious example might be flowing water -- when was the last time anyone saw a reverse waterfall, with the bits of water all jumping just right to flow uphill, for instance? No doubt this has to do with the low probabilistic chances of things coming together in just that way; unfortunately, probability and statistics is a subject that has a special "confusion potential" all its' own! Earlier, on the "Computability of Consciousness" and "Qualia" threads, I talked about the proper calculation of statistical things like *entropy* being dependent on the knowledge of the observer doing the calculating, with Nature hopefully being predictable, in the sense of not biasing the statistical outcomes in some really strange or unexpected way from the point of view of any physically reasonable observers. Apparently, we have to assume that the universe is not too disordered or too scrambled to begin with, with at least the potential for some observers or measuring devices to gather some information without melting down or something. That way, you get a dynamic arrow of time even if the underlying forces are symmetrical.
>
> Using the term "causality" to refer to what I would call
> "monodirectional causality" doesn't mean that the laws of physics agree
> with you. Nor does a deliberately narrow definition of "causality" mean
> that all order and causality get tossed out the window if the definition
> turns out to be wrong.
Well, in Many Worlds theory, there apparently has to be a dynamic time arrow in the sense that events tend to cause alternate timelines to separate or "decohere" from each other, as time moves forward into the future. When you think about it, it would be really weird, right, if the separation went the opposite way, i.e., if we all had alternate pasts, and were forever digging up authenticated histories of past events that thoroughly contracted one another? So the use of Many Worlds to get rid of Grandfather Paradoxes seems consistent with the fact that Niagra Falls flows downhill?
David Blenkinsop <blenl@sk.sympatico.ca>