Wednesday, September 26, 2012

Meaningless Conflict


Boudry and Plantinga argue about the bit that God did or didn’t do.


My favourite atheist, Larry Moran, has once again alerted me to some very interesting material that warrants a blog post. In a post entitled Boundry vs. Plantinga and dated September 20th Larry gives us news of a review by Maarten Boudry of Alvin Plantinga’s book Where the Conflict Really Lies. This review can be found in the IHPST newsletter here. (Just search for “Boudry”). Needless to say atheist Boudry pretty much rubbishes Plantinga who is Christian. For myself, however, I find I am unable to support either side of this debate because both parties seem to be deploying some very dubious concepts and arguments. To be fair I haven’t read Plantinga and I am only going on Boudry’s representations, so it could be all Boudry’s fault. Below I have reproduced the section of the review I’m going to consider here:

In Part I, Plantinga argues that theism is not in conflict with evolutionary science, only with unguided evolution, which is a metaphysical add-on, not a part of evolutionary theory proper. But what about the thesis that the source of mutations is random, an assumption that is part and parcel of evolutionary theory? “Random” in this context does not mean pure chance, but rather without foresight, or not necessarily concordant with the organism’s adaptive needs. According to Plantinga, random mutation is compatible with God orchestrating the whole process. It is just God’s way of creating novelty. But there is no wedge to drive between “random” and “unguided”. If God is steering evolution in the right direction, making sure that suitable mutations arise, the mutations cannot be “random” in the technical sense, for then they would occur for the benefit of the organism (or whichever plan is on God’s mind). This becomes clear when Plantinga, strangely enough, endorses creationist Michael Behe’s argument from “irreducible complexity” (IC, see further), a pseudoscientific concept that has been completely rejected by the biological community. Behe’s argument is that the evolution of IC systems requires the simultaneous occurrence of different beneficial mutations, which is exceedingly unlikely. Therefore, some intelligent designer must have monitored the development of IC systems, if not by miraculous creation, then at least by causing the requisite mutations for natural selection to work on. But then, obviously, the mutations arise to meet the organism’s adaptive needs, and are no longer “random”. That is exactly the point of Behe’s argument. Besides, why doesn’t God skip the cruel selection part and create a whole population of well-adapted organisms without further ado? (See below for Plantinga’s answer to the problem of evil.)

The thesis that mutations are random (i.e., unguided), rather than being a metaphysical afterthought, has been amply demonstrated and is accepted as the null hypothesis by evolutionary biologists. Experiment after experiment has shown that there is no evidence of nonrandom mutations arising because an organism “needs” them. Further, if some intelligent agent is triggering mutations after all, it seems that he/she/it is causing precisely the kind and rate of mutations that one would expect if the process were entirely undirected. Plantinga’s effort to stave off the conflict between theism and evolution is a failure. Either he is buying into creationist fantasies that have been put to rest long ago, or he is hammering on the excessively weak claim that it is logically and metaphysically possible that, all evidence to the contrary, evolution unfolds under supernatural guidance. But if the bar for rational belief is lowered to mere logical possibility, and the demand for positive evidence dropped, then no holds are barred. Evolution (or gravity, plate tectonics, lightning, for that matter) could as well be directed by space aliens, Zeus or the flying spaghetti monster. (I was going to  include the devil in the list, but then it turns out that, on page 59, Plantinga has no qualms at all about treating the horned one as a serious explanation. There goes my reductio.)

In the above I see no cognizance of the single most important factor which makes or breaks standard evolutionary theory*, namely the critical importance of the physical regime in making the evolution of organic configurations possible. It is this regime which determines the size and shape of the space of permissible configurations in which random thermodynamic shufflings are taking place. If this space is wrongly shaped or too large then the thermodynamic agitations which effectively constitute the computational activity which “seeks out” stable organic configurations, is easily overwhelmed. But if our physical regime actually returns a sufficiently constrained shape and size to configuration space then the evolution of organic forms may have a realistic probability.

If for the sake of argument we assume that the latter is the case then no further “guidance” (whatever that means) is needed for evolution; there is then sufficient mathematical providence for evolution to proceed. For if the configuration space set up by our physical regime is sufficiently constricted then the random thermodynamic shufflings become the computational searching by which the configurations of life are located and locked into place by the evolutionary ratchet.

The “shuffling” patterns of randomness (constituting the natural engine of computation) are no more and no less than that; namely, a particular class of mathematical patterning. How then are we to distinguish between patterns that are guided (Plantinga) and patterns that are unguided (Boudry)? What criterion do we use to distinguish between guided and unguided patterns? To me the dispute between Boudry and Plantinga about the nature of evolution seems to be Unintelligible. Unless, of course, it is claimed that our physical regime is insufficiently constrained to impose the right shape and size on configuration space – in which case the evolution of life would require something extra – perhaps some unknown physics or even the enigmatic “tinkerings” of a homunculus intelligence working behind the scenes, not unlike Maxwell’s demon! If so then Boudry and Plantinga would have a meaningful dispute. Otherwise I see little content in their argument.

However, if the mechanisms of evolution (and OOL) are as the evolutionists think they are then I agree with Boudry that the random shufflings of the computational engine are going to be statistically independent of an organism’s need to survive: Unlike Maxwell’s demon those shufflings don’t have any information about the direction in which they should take; in fact that’s precisely why a search needs to be conducted! But I disagree with Boudry when he says that Irreducible Complexity is a pseudoscientific concept. You see, if Boudry accepts the established view of evolutionary mechanisms then he is implicitly accepting that configuration space has been arranged with the right shape and size needed to confer upon those mechanisms a realistic probability of generating organic forms. This shape requires stable organic structures to be close enough together in configuration space to form a linked set so that evolutionary drift can proceed in a series of incremental steps; where each step is a small jump to the next relatively stable organic form. This condition whereby organic structures form a linked set in configuration space is what may be termed “Reducible Complexity”. But in positing this concept we beg the question of whether or not it is in fact the case: What if organic configurations do not form this connected set? What if organic forms are isolated islands of stability in a huge sea of unstable configurations? Irreducible Complexity as a concept is simply the negation of reducible complexity and as such it is both a meaningful idea and the subject of legitimate scientific inquiry. Should irreducible complexity actually be the case then needless to say the established view of evolutionary mechanisms would simply not work and the posited engine of natural history would have to be fundamentally revised. Perhaps this may explain why Boudry shrugs off IC in an offhand way, rather than face the uncharted intellectual vista it brings into view. He is using the word “pseudoscience” as the establishment once used the cry of “heresy”.

If engines of evolution and OOL are as the scientific establishment would very likely prefer them, then this entails  the supply of some very particular mathematical conditions: These conditions must be constantly maintained as the status quo of our physical constitution and therefore these conditions are arguably an ongoing form of guidance, although guidance that is rather different from the default “interventional” theology that envisages some kind of homunculus or "demon"; a model that is affirmed by many theists and vehemently attacked by atheists like Boudry.

Boudry’s review naturally runs into other issues like the logical possibility of miracles, the problem of suffering and the destructive self-referencing of naturalism. I may look at these in a later post.


Footnote
*In this context I’m bundling organic evolution and OOL into one package and calling it “evolution”.
.

Wednesday, September 12, 2012

Hard and Soft Science in the Junk DNA Debate

....and sociology, history and theology books are even more comfortable!
Atheist Biochemist Larry Moran will no doubt think of me as a nincompoop for entertaining theism. Nevertheless, although he is a very dogmatic atheist I have found him generally polite in dealing with theists like myself. I may disagree with Larry’s vision of science and find him a bit of a curmudgeon, but on balance I would describe him as a gentleman and scholar and by and large fair in his comments (If you are on the right side of him!). In any case I’d far rather leave comments on his blog than the blog of PZ Myers, a blog which is populated by some very rough customers indeed, not to mention Myers himself who lumps together all theists for his uncompromising contempt. His blog is more like a piranha invested river and when the fish there are not ripping to shreds hapless Godbots they start tearing at one another. In comparison Larry’s blog seems a much more civilised environment.

Perhaps the foregoing explains why I surprised myself in feeling a bit sorry for Larry when I found him heart aching over the latest news from the ENCODE project which is being touted (or “spun” according to Larry) as evidence that most (~ 80% ?)  of the human genome has some kind of function after all. Larry has put a huge intellectual stake in the possibility that the genome carries a lot of redundant information – that is, information that has little or no function. This probably follows from his view that evolution is primarily a random trial and error process, a process that is likely to leave in its wake an untidy backwash of redundant configurations. To Larry, cosmic disorder is a sure sign of a lack of design and therefore absence of evidence of God, or for that matter any other kind of intelligent designer. Larry atheism’s is bound up with questions over the nature of randomness; in particular the philosophical question of whether or not we should consider it unremarkable and attribute to it a null computational complexity.

Now I’ll admit that I’m not much of a biologist or a sociologist (two subjects I consider to be related in a broad inter-disciplinary spectrum). My background is in physics and programming and this has probably  left me with an intellectual impediment; it means that when I hear the term “prediction” I think of it in very mechanical terms: I imagine a theoretical mathematical/algorithmic system that is tentatively put forward as isomorphic in some way with the ontology of our world. One then cranks the handle of this mechanical system and it churns out fairly clear cut “predictions” (Although sometimes with a probability attached). If the predictions fail then because the theoretical narratives of the “hard” sciences are relatively simple affairs with few adjustable variables then theoreticians don’t have much maneuver room to “get out of that!”.

Needless to say in the soft sciences this is much less the case. The soft sciences deal with highly complex objects that may comprise of an open ended set of impinging variables. Consequently “predictions”, so called, are far less unequivocal: A failed prediction need not necessarily be fatal as it may be possible to propose not just a variable adjustment, but perhaps hitherto unknown variables in order to accommodate an unexpected result.*

On balance evolutionary theory is, to my mind, a soft(ish) science, and this is particularly clear with evolutionary psychology, a “discipline” that generates great contention; a sure sign that people’s retrospectively applied imaginations are hard at work here, untrammeled by the dispassionate tester and arbiter of actual involuntary experience. Similarly, I can’t see how the absence or presence of junk DNA is a hard prediction of evolutionary theory. In fact Larry quotes another non-biologist who is now trying to make evolutionary sense, post-hoc, of the claimed absence of Junk DNA:

So they assumed it [Junk DNA] was left over from evolution, had no current function, and was, literally, junk. As Francis Crick, one of the Nobel Prize winners for helping discover the structure of DNA, put it, non-coding DNA has “little specificity and conveys little or no selective advantage to the organism”. Right. As though nature would waste that much energy.

Here we have a evidence that evolutionary theory can very readily assume a retrospective sense making role; it’s not that there’s anything wrong with that because in the absence of a capability to arrive at hard predictions embedding post-facto data in an imaginative theoretical narrative may be the only way we can make (controversial) progress.

I don’t think anyone is going to be surprised to hear that just at the moment the Homunculus ID community are making hay while the sun is shining. They have long maintained that their version of ID “predicts” (Haha!) that most of the genome is functional. To them the “spin” surrounding the ENCODE project is a great vindication.  But the irony is that to my mind the absence of Junk DNA is hardly a hard prediction of even Homunculus ID, as I shall now explain....

Evolutionary theory may be a softish science, but in my view Intelligent Design theory is a lot softer. As I made clear in my series on ID predictions it is difficult to get any hard predictions out of ID at all. Intelligence is in fact synonymous with personality and personality is a blend of emotion (another word for motivation) and the ability to solve computational problems. Necessarily intelligence carries with it a large measure of inscrutability about its motives and methods. Because, then, personality is by definition a highly complex object with many hidden variables the only way I can see of getting predictions out of ID is to start making assumptions about the nature of the intelligence one is dealing with; in particular one might assume that the intelligence which designed life has similarities with a human molecular engineer and one can then apply a whole array of observations on technological (wo)man to anticipate how an alien molecular engineer (or “Homunculus”) might design life. The upshot was that I could by and large agree, after a fashion, with most of the predictions that appeared on the Uncommon Descent post that was the subject of my ID predictions series. But the irony here is that the absence of Junk DNA was one of the few predictions I just couldn’t confirm as an outcome of Homunculus ID. It seemed to me that it was a toss-up as to whether an alien molecular engineer may or may not leave “commented out” information in the genome about his essays in biology.

Summarising: It seems that the junk DNA debate is in a position where the respective sides are investing themselves in conclusions that as far as I can see are not unequivocal predictions of their respective logic. The trouble is the entrenchment is now too far gone for the two sides to easily back out..

It is clear that the homunculus assumption is by no means sound: After all if we are ultimately dealing with the kind of God envisaged by Christianity then we are dealing with something of an entirely different genus to a tinkering alien or quasi-human intelligence. God is an immersive totalizing reality, the entity all things find themselves to be immersed “in”. (Acts 17:27-29 ). As a fellow Christian once said to me: It is as if we are a thought in the mind of God and as such need constant attention and maintenance as thoughts do or else conceivably they could be snuffed out at any time. That idea has not just stayed with me but was reinforced when I started probing the nature of intelligence and found it to be an exploratory searching process: In fact the cosmos is shot through with the “search-test-reject-and-select” paradigm. As has been remarked before, the universe does look more like a giant thought than a giant piece of hardware tinkered with by an alien intelligence.

The pervasive nature of God in most conventional theologies is echoed in our cosmos, a cosmos necessarily pervaded by a grand logical hiatus. That hiatus is everywhere and everywhen. This pervading hiatus invites imaginative post-hoc theologies in order to make sense of it. But pervasive facts of life, like the peculiarity of there being something rather than nothing, are easily taken for granted as somehow intrinsically necessary in and of themselves: To a creature like (wo)man "programmed" to take note of changes in the status quo total immersion phenomena may be filtered out as not interesting.  Consequently theologies are always going to be controversial.


Footnote
* We are really talking about a spectrum of objects here – there is no is sharp demarcation between the objects of the soft sciences and those of the hard sciences.  Hard science imperceptibly merges into soft science.