Tuesday, May 23, 2017

Omniscience and Omnipotence in Evolution 2.0


The word 'omniscience' seems to play an important role in Perry Marshall's characterization of the Intelligent Design research program as a specie of Old Earth Creationism.  In at least two places (here and here) Marshall seems to claim that Stephen Meyer has brought in omniscience into their debate on Unbelievable.  What Meyer actually said (twice) was that Marshall seemed to be attributing a "near omniscient" knowledge to various kinds of cells.   

Before teasing apart what this means, let's look at one of Marshall's criticisms of neo-Darwinistic evolutionary theory:
Darwinists have this naive conception that random copying errors with natural selection have godlike powers and can do anything.
Of course this violates every principle of information theory and probability known to man; this is why Darwinists scream bloody murder any time someone brings up statistics. But it’s impossible for accidental processes to do this. They may as well believe in Superman or the Easter Bunny.
Marshall claims here that neo-Darwinism attributes to mutation+selection a near-omnipotence if not omnipotence.  However, the sort of omnipotence that neo-Darwinism claims for mutation+selection (+ even more accidental mechanisms such as genetic drift) is just the power to get from a prokaryotic cell even more simple than a bacterium to every "leaf" on the evolutionary tree.  But this is the same sort of power that Marshall claims for a "Swiss Army Knife" of mechanisms that are "in principle, capable of getting you [from] anywhere on the tree of life to anywhere else." Not truly omnipotent per se, but presumably capable of accounting for anything we've ever seen.


The reason that Marshall feels comfortable claiming this quasi-omnipotence for his "Swiss Army Knife" is because it is evidence of automated process that has intelligence somehow built into it (as does any AI or expert system), and that sounds so much more likely than the lucky mutation theory that has been "settled science" for the last 80 years.  It seems possible because, presumably, the evolving cell has such great knowledge about itself that it can not only rearrange its genetic elements responsibly but can tell what sort of changes will likely help itself survive in a given environment because of its purported understanding of both its genetic architecture and its environment.

Ultimately, this sort of quasi-omnipotence comes from having very specific knowledge and a  powerful, perspicacious application of that knowledge, something that is somewhat oracular in nature, where a species just always figures out what it needs to get from "anywhere on the tree of life to anywhere else."

Marshall is at least 10% sure that abiogenesis will be solved in our lifetime in spite of the fact that for unguided evolution to work (I'll use "unguided" for now, since Marshall has hijacked the word "accidental" for his purposes), these mechanisms must evolve in cells that don't begin being able to use them.  So how does someone acquire superpowers?  It's very simple.  That someone must first acquire the ability to acquire superpowers!  If that person can first acquire the ability to acquire the superpowers through a natural, repeatable process, then I no longer have to claim that the acquisition of superpowers is 'accidental', because now he can get whatever superpowers he needs for a situation.  Once again we've saved naturalism through the sheer power of speculation.  But the only (near-)omnipotence on display here is the power of speculation that underlies evolutionary logic.   

Some polyploidy does occur naturally in plants. It exists even more commonly in human-cultivated varieties, because farmers have worked hard at making the most of the accidents.  In the animals, there is a sort of modest polyploidy with regard to hox genes in the sense that there are 4 sets of these to the invertebrates .  There is a reason that we don't currently see invertebrates using polyploid hybridization .  What is the reason for that?  And why does that reason not apply to the remote history about which evolutionary speculates?  So there is a certain static homology between the sets of Hox genes in vertebrates that is being attributed to a polyploid hybridization event.  Plants do it, so why not animals? That question is not likely to be raised (since this seems to kill all animal embryos), but if it is you can bet it will only be for handwaving on why animal polyploidy was possible 500 million years ago it the Cambrian orgy of innovation and not now (the explanation probably involving "canalization" or "irremedial complexity" or the like).

Still, polyploidy doesn't explain much for animals (we can't repeat the supposed event) and explains even less for plants.  Broccoli is a polyploid hybrid of two closely related cabbage species.  The species are different enough to not hybridize well in the normal way, but can hybridize fine by getting a complete set of chromosomes from each parent.  Polyploidy is less providential but more limited than horizontal gene transfer, but both mechanisms involve consolidating pre-existing solutions in new ways; they don't explain the origin of novelties any more than accidental mutation.  A lot of mutation is not 'accidental' sensu Marshall, but we also don't have any good causal explanations why that is.  Directed mutation is yet another tool in the microevolutionary toolkit that neither explains itself nor other tools nor the origin of evolutionary novelties.  

The only potentially creative mechanism in the 'Swiss Army Knife' toolkit is the ill-defined concept of James Shapiro's "natural genetic engineering," impressive mechanisms that are poorly understood (why don't more cells do this DNA shuffling trick that some protozoans do?).  We don't understand the limitations of these mechanisms, so evolution-of-the-gaps means we must attribute omnipotence to this mechanism whatever it turns out to be.  But let's not characterize the oracular nature of what Marshall sometimes describes as "algorithmic" mechanism as "near omniscient."  Marshall hears 'omniscient' and goes on a god-of-the-gaps tirade, when he is attributing a sort of 'omnipotence' to the intrinsic nature of cells' ability to discover and understand, and is effectively practicing a cell-intelligence-of-the-gaps.  Since science can't determine (according to Marshall) how a designer (who wouldn't have to be omniscient according to Marshall since omniscience isn't required) could have gamed the process, the cell must be presumed to have figured out macroevolution through its "savant-like genius." 

In order to explain a macro-evolutionary process (on the grand, omnipotent scale of "unlimited variation") that requires no recourse to a designing mind outside the process (i.e. the causal powers of the process are well established), we actually need the accidental creation of cybernetic metabolic systems with the power to direct their own mutation and change.  The theory for that creation does not exist in the over-the-top sense that Eugenie Scott attributes to the word "theory."   Scientists should have the freedom to develop such a theory, even if that means lots of speculation at this point.  But aside from understanding the sophisticated mechanisms of microevolution that life is capable of (and that even creationists typically attribute to the natural order) through real experiments (which, like Marshall, I'm glad biologists are finally doing after all this time), all there is is speculation about the history of life on earth.

Meyer hasn't snuck in omniscience into the debate any more than Marshall has snuck in omnipotence.  Having dismissed neo-Darwinism as believing attributing "super powers" to natural selection and mutation like believing that a flea ca "leap tall buildings in a single bound," Marshall points to a guy with a hangglider and attributes the powers of Superman to him instead.  While it is superficially more plausible that the guy in the hang-glider has the powers of Superman than a flea (he certainly stays up in the air longer than a frog does), the glider can't fly around the world at Mach 9.  It's an impressive tool in the hands of a skilled flyer, but its wonder-working power is very limited.  

No comments:

Post a Comment