Jason Cawley
Wolfram Science Group
Phoenix, AZ USA
Registered: Aug 2003
Posts: 712 |
Point of correction (to a disrupter, since removed) - you said at one point "as Wolfram has pointed out, the real goal of evolution isn't so much 'survival', as it is diversity." I am quite sure you will find he has said no such thing. Don't put words in his mouth.
Complexity is a real phenomenon and as such needs an explanation. It is not a "goal". What Wolfram has said about evolution - which happens and matters obviously, and involves natural selection on the basis of survival, not any other "goal" - is merely that it is not obviously the only explanation of complexity within life. Notice, complexity is not everything in life. Nor is it restricted to life. Complexity is a more general phenomena, it happens in systems that are not alive at all, and for essentially algorithmic reasons - the typical behavior of simple programs beyond a modest level of internal variety.
As such, there is no necessity that a selective advantage causes a given pigmentation pattern on a sea shell (just because the latter is complicated), when there clearly isn't such a cause for a drainage network or a pattern created by fracture. Complex patterns can arise for essentially computational reasons in life and outside of life. They then no doubt feature as inputs to natural selection processes, as any other source of variation can (mutation, genetic mixing, drift, etc).
The point is merely that algorithmic variation can enter at that pre-selection stage, as one kind of "variation" in the formula, "descent with variation". And can be entirely adequate to cause significant complexity, on its own. The point is not that complexity is a goal of anything, whether of rule 30, or the edge of a broken stick, or the billows of a cloud. Just that it already happens in these prebiotic cases and can be expected to continue doing so in organic ones. Natural selection then undoubtedly fiddles with the resulting varied and complex output - but it does not need to produce them in the first place. The nature of simple programs will do that.
This has testable implications, which can distinguish which basic cause is operating. If a full range of parameters for a complex subsystem is seen empirically, the inference is that selection has not picked special parameters to produce that variety. If only a few optimized special cases are seen, natural selection is the likely explanation. If a parameter space is seen to be "winnowed" or "culled", with portions remaining but others not present, the inference is underlying variety pruned by natural selection.
The timing of variety and optimization can also give evidence for or against this proposition. Wide variety quickly and first suggests an algorithmic "variety motor". Elaborate "engineering" optimization and specificity, on the other hand, suggests natural selection. Both can be presumed to operate, and which is more important in a given case is an empirical question about what actually happened, not an a priori one.
Report this post to a moderator | IP: Logged
|