Evolution and Islands of functionality

I've said it before and I'll say it again: William Dembski, the North American "Intelligent Design"  guru, is a nice bloke and in many ways an admirable Christian; moreover, I think one of his primary publicized conclusion is entirely correct; that is, a universe such as ours, especially given the presence of life, demands a huge upfront information input. Unless we are going to invoke multiverse ideas this is a truism whether or not life is a product of the mechanisms of evolution as conventionally conceived (But see here for qualification). Dembski is also a reasonable Christian who disowns the fundamentalism abroad among many US Christians. But in spite of all this he has been rejected and even abused among some evolutionists of the academic establishment, especially by evangelical atheists. This is at least in part because some in the IDist community have assumed his work is a sure fire refutation of standard evolutionary mechanisms. But Dembski's main conclusion isn't such a refutation. In fact Dembski has given a back-handed acknowledgement of this fact. 

As I described in my last blog post there are big stakes here as a consequence of the US right-wing IDists and the atheists in the academic establishment polarizing around what they both believe to be a sharp dichotomy between "natural forces" and "intelligent agency". But the neutrality of Dembski's initial conclusions doesn't mean that Dembski is what the IDists contemptuously refer to as a "Darwinist"; rather he very much aligns with the IDist community and argues against standard evolutionary mechanisms as we shall see in this post. 

Given the establishment vs popularist right-wing polarisation in the US, it is not surprising if Dembski has been embraced by the right-wing and he has turned his talents toward supporting some of their contentions. For example in this blog post of his we find him entertaining (but falling short of outright affirmation of) the theory that Covid 19 was genetically engineered in China. His post will go down well among Trump right-wingers. In fact I'd be interested to know whether or not Dembski is a Trump supporter and believes in a stolen election. 

For myself I have no useful input on theory that Covid 19 was genetically engineered in a Chinese laboratory and then perhaps accidentally released. It is a plausible theory that may or may not be true as far as my knowledge is concerned. Unfortunately the authoritarian and secretive  nature of the Chinese regime doesn't help their case one little bit: It would be typical of a totalitarian government with little or no accountability to host a classic cock-up and cover up scenario like a laboratory escape. But if Covid 19 is a Chinese contrivance I think it unlikely it was deliberately released; that idea just smacks too much of the cold hearted Machiavellian fantasies spread about by the deluded conspiracy theorists; I find incompetence and cover up scenarios much more plausible and in line with humanity's often sleazy and idiotic behavior. In any case it cuts both ways; that lab-leak theories serve right-wing tribal interests erodes the credibility of these theories. But I'm less interested in this issue than Dembski's references to the evolution question.

***

So, as I was saying, Dembski's main work doesn't contradict standard evolution:  But even so, as I've said, Dembski, of course, finds himself on the anti-evolution side of the culture war and naturally enough has tried to advance arguments which attempt to refute evolution. In his Covid 19 post he does a resume of a frequent argument used by IDists. In his post we find the picture I've published at the head of this post and Dembski tell us about it:

This first slide illustrates, by analogy, what the Darwinist thinks must be the case, namely, that islands of functionality exist dotted along the way in getting from the left most island to the farthest off island. With all these intermediate islands, it is easy (probable) to jump from one island to the next and thus get to the far-off island by starting with the closest one (the far-off island representing the end product of evolution that we’re trying to explain).

Yes I agree, each organic variation that walks the Earth must be functional and able to transmit incremental variations to the next generation that themselves must also be functional.  Evolution is a step by step gradual process that doesn't conceive huge organic variations appearing in one generation. e.g. Lobe finned fish didn't become amphibians in just one generation; that would require millions of years of step by step change, where each step is capable of survival and replication. 

But Demsbki goes on to give us this second picture to ponder.....: 

According to Dembski this picture illustrates the possible problem with standard evolutionary mechanisms that depend on the small jumps of incremental change. Of this matter he says this:

But how do we know that those intermediate islands exist? The second slide illustrates this possibility, and insofar as it describes what is happening with biological change, it renders Darwinian evolution far less plausible. It needs to be noted here that whether these transitional islands (i.e., intermediate functional biological forms) exist is a matter for fact. The dispute between design theorists and Darwinists is over the evidence for these intermediate islands/forms. For the Darwinists, these intermediates must exist because Darwinism requires a gradual form of evolution. For the design theorists it’s not that these intermediates can’t exist but that they might not exist and if they don’t, that argues for intelligent design.

Yes again I agree: For the Darwinists, these intermediates must exist because Darwinism requires a gradual form of evolution. The battle between IDists of Demsbski's variety and the establishment evolutionists revolves round the attempts on the one-hand of IDists to show that there is no evidence for this "island" hopping scenario and on the other hand evolutionists trying to show that there is evidence of the existence of closely set islands of functionality.  The IDists, of course, are quite sure that islands of functionality are not closely set enough to facilitate evolution and they then invoke their so-called "explanatory filter" and out pops intelligent agency (I believe to this explanatory filter to be flawed if pushed beyond everyday application into the realms of the origins of life - see here for more details). 

***

But  there is one thing that Dembski's island metaphor hasn't made sufficiently explicit in my opinion. In the first picture above it could be that the sea is actually very thickly populated with islands of functionality and that the distance between these islands is a small configurational step. And yet this in itself, although a necessary condition for evolution, isn't a sufficient condition. This is because the islands may be so small that a random hop has very little chance of landing on any of these tiny islands of functionality. Actually, if one blows up the magnification of this "many small islands" picture it starts to look a little like Dembski's second picture with islands well separated. In fact it's vaguely reminiscent of what one sees of a galaxy in space - from a distance they look to be crowded with closely set stars - but blow up the magnification and one finds the stars to be very small and too far apart for space travel. Likewise, there may well be many islands of functionality and not very distant from one another in terms of steps but because they occupy such a small area in the "sea of non-functionality" random island hopping is too improbable to be practical.

One way of thinking about this situation is to understand that organisms, because they are composed of many particles, are actually multidimensional entities with huge numbers of dimensions. There may be many functional configurations within a few short steps but nevertheless too few, given the number of dimensions, to be accessible with small random hops; the overwhelming number of short hops will go in the wrong direction.

I actually much prefer what I call the "spongeam" picture to Dembski's first figure above. I have featured the spongeam structure on this blog several times before. It looks something like this:

In this metaphor we are in 3D rather than 2D, although of course we should be talking about a configuration space of immense dimensionality and where the spongeam structure is considerably more tenuous looking than it looks in the picture above.  However, the spongeam metaphor, in my opinion, conveys, the complexity of the situation better than the island picture. In the spongeam picture I identify the necessary condition for standard evolutionary mechanisms to be that the class of functional, self perpetuating organisms form a connected set in configuration space, resulting in a thin, tenuous, but complex network of fibrils spanning a space of immense dimensionality.  In this picture the random walk steps of evolution are modeled as a form of diffusion guided by the thin connections (or channels) of the spongeam. If the spongeam exists then the mechanism of evolution is a process of diffusion through this network of channels. Also, as I've remarked before, one can express this metaphor for evolution mathematically. Viz: 

I explain this equation more fully in this blog post.  Suffice to say here that Y represents some kind of population density at a point in configuration space. The first term on the righthand side is a diffusion term resulting of the random hops across the space. The second term on the righthand side represents a breeding or decaying population term, where V is a value which varies across configuration space. It is this value which describes the spongeam structure, a structure which must be sufficiently connected to provide the necessary conditions for standard evolutionary mechanisms. It embeds the upfront "Dembski information" required for those mechanisms to work.

Like Dembski, I have doubts that this necessary condition is actually fulfilled given our current understanding of the physical regime in spite of the stringent constraint that the known laws of physics put on the possible behaviours in configuration space. My feeling is, and I admit it's only a intuition, that the high organisation of life means that the number of possible organic structures are likely to be overwhelmed by the number of possible disordered configurations. That is, notwithstanding the known laws of physics which considerably reduce the "volume" of configuration space, there simply aren't enough viable organic configurations to populate configuration space with an extensive connected structure like the spongeam, a structure which is a necessary condition for molecules to man evolution. So, it may be that IDists like Dembski are actually right. But having said that I don't think the case against evolution is actually proved and standard evolutionary mechanism may yet be the engine driving natural history. I'm not strongly aligned on this question.

It is ironic that in one sense IDists of Dembski's ilk would likely agree with the academic establishment on one very important aspect of evolution; namely, that the fossil record testifies to a natural history of changing life forms over millions of years; So, in the natural history sense they both accept that evolution has occurred although would disagree on the underlying driving mechanisms: A further irony here is that the mechanisms of evolution, when stated in their most general form, even by an evangelical atheist biochemist like Larry Moran, admits intelligent design as a possible driving mechanism - see here. I wonder if Moran is aware of this? 

So, at heart the contention between IDists and establishment evolutionists is about the nature of the internal engine driving evolutionary change. But I have doubts that this contention can ever be settled conclusively given that fossil, genetic & breeding data can only ever be a way of sampling the highly complex processes of natural history. I'll have to leave the two sides arguing the evidence for that one, although I'm inclined to float my vote against the spongeam as a reality and yet at the same time stand with the evolutionists in the culture war against an extreme right-wingery which of late has manifested itself as a threat to Western democracy.

***

If the class of functional structures are closely spaced but occupy too small a "volume" in configuration space to be reachable by evolutionary diffusion, there may yet be a way round this situation, one that I've probed for many years (although without unambiguous success). If some kind of tentative expanding parallelism is in operation which probes a few steps across configuration space these islands of functionality could then be reached. But accompanying this there would have to be some underlying drive to preferentially select these islands and that aspect, which implies a built-in teleology in the physical regime, would have to exist. Quantum mechanics gives us expanding parallelism straight away; it also gives us the selection too, in the form of collapse of the wavefunction (I'm by-passing multiverse & decoherence interpretations of quantum mechanics here). But apparently (and I stress apparently) quantum selection isn't preferentially biased but random. - as far as we know. Notwithstanding that however, my radical suggestion is that there is an underlying teleology in the cosmos, a teleology embodied in a biased seek, inspect, reject and select algorithm behind the generation of life. In effect  this would both considerably speed up the diffusion and introduce a life favoring value of V in the above equation. 

Well, I suppose it's likely I'm on a hiding to nowhere here, but in the poisonous atmosphere of a  polarised culture-war, to even tentatively investigate such ideas is an affront to the hardened nihilistic atheists and a heresy to the hardened dualists among right-wing IDists & fundamentalists. Just as well I'm in a relatively unconnected domain on this part of the web; I'm not keen on meeting them. (I've already had three unpleasant chance web-meetings with fundamentalists and/or conspiracy theorists - see Richard Sweet,  Steve Pastry and Ken Ham)

For evangelical atheists whose world is ultimately meaningless these ideas would smack too much of intelligent contrivance & purpose to be acceptable. Take an atheist like Steven Weinberg for whom the universe is absurd: As his famous saying goes "The more the universe seems comprehensible, the more it seems pointless".  And yet I can empathize to some extent with Weinberg: From a human perspective science and industry have left us with enigmas & challenges to deal with. But even so perhaps Weinberg could have asked himself more questions about the origins of a comprehensible cosmic order rather than jumping to the conclusion that it's all absurd.

It is ironic that for Christian IDists and fundamentalists the thought of a cosmos provisioned to generate life via a teleological version of evolution is, if anything, an even greater affront than it is to nihilistic atheists; the latter will brush you off as a fool, whereas right-wing religionists are inclined to see you as a subversive, may be even a malign & wicked influence. For one thing they have difficulty with the notion that God's creation is able to create of information. But creating information is what teleological algorithms achieve. There is nothing intrinsically anti-Christian in seeing human beings as a thoroughly "natural" (sic) phenomenon; for as far as we know they are a dynamical pattern that works within the operational envelope of a physical regime that the sovereign Creator has set up and manages on a moment by moment basis. In that sense human beings are at once both natural and supernatural. Moreover, as a "natural" phenomenon human beings (like natural history itself) are clearly able to create information on a daily basis. But the dualistic religionists who have committed themselves to the notion that intelligence is tantamount to some form of mysterious intellectual alchemy that cannot be described in algorithmic and material terms have backed themselves into a corner: Their vested interest in a particular line of thought brings down a taboo on any suggestion that in God's "natural" (sic) world information is being created all the time. Why is this such a difficult idea given that God is sovereign and it is God's created world? The temptations of gnosticism are never far away.