Evolution: Naked Chance?

A set of basic construction parts; throw in a few gluons (i.e. nuts and bolts) and then agitate randomly. Will this mix generate self-replicating, self-perpetuating configurations? There is a way of doing it but is this The Way it happened in our cosmos? (Probably not!)

This post is a response to biochemist Larry Moran's essay entitled "Evolution by Accident".  In his essay he leans toward the view that randomness  and serendipity play a big very big role in evolution. He sets himself against the natural selectionists whom he characterises as supporting a "non-random" view of evolution. In his conclusion he writes:

I've tried to summarize all of the random and accidental things that can happen during evolution. Mutations are chance events. Random genetic drift is, of course, random. Accidents and contingency abound in the history of life. All this means that the tape of life will never replay the same way. Chance events affect speciation. All these things seem obvious. So, what's the problem?

The "problem" is that writers like Richard Dawkins have made such a big deal about the non-randomness of natural selection that they risk throwing out the baby with the bathwater. A superficial reading of any Dawkins' book would lead you to the conclusion that evolution is an algorithmic process and that chance and accident have been banished. That's not exactly what he says but it sure is the dominant impression you take away from his work.

Here's another example of the apotheosis of chance in Moran's writings:

What about Monod's argument that evolution is pure chance because mutations are random? Doesn't this mean that the end result of evolution is largely due to those mutations that just happened to occur?

There is no need for me to take sides in this debate between evolution by natural selection and evolution by "pure chance". In fact for the sake of my argument I could proceed under any mix of natural selection and so-called "pure chance".  What I want to show here is that yes, current notions of what drives evolution entail a big random factor, but it is only one aspect of evolution and there are other aspects which are even more significant.

My point will be this: Evolution is driven by chance, but it certainly isn't naked chance; in fact overall the process is very, very far from naked chance. But reading Moran's essay one could be forgiven for thinking he's pushing naked chance too far and has no awareness of how ordered, a priori, the world must really be for evolution to work. The second quote above from Moran would be less misleading about the process of evolution if "pure chance" were replaced by "chance". Evolution cannot be pure chance as we shall see. Moran is either oblivious to this fact or sees it as not worthy of note; but I see it as highly significant, in fact evolution's most significant feature.  The fact of the matter is that the chances in the process of evolution must be subject to a highly constrained envelope if it is to work.

This apparent obliviousness of Moran to the organising envelopes which must constrain the chance diffusion of evolution may trace back to Moran's very partisan form of atheism and the related sentiments I identify in my post on the many worlds cosmology: The purposeless backdrop of the atheist world-view has grave difficulties in making sense of any built-in cosmic bias or preference toward certain states of affairs; in particular an ordered status quo. This bias leads to tricky questions like why this and not that?  These questions in turn may raise what to some is the demon of "purpose" and from there it is short walk into at least a conjectured theism. Partisan atheism finds any contingent bias in the cosmos uncomfortable and finds it easier to handle a cosmos where everything is evenly favoured with the "butter" of probability spread uniformly, thus betraying no sense of preferred statuses in the cosmic dynamic (See here).

My arguments follow even if natural history is driven by some mechanism completely different to standard evolutionary mechanisms. This is because my argument is about logic; that is, it is of the kind "If so & so then it follows that......etc".  Where "so & so" stands in for "standard evolution" which may not actually hold good (and I must confess I have my doubts!). What concerns me here, however, is what follows if we assume standard views of evolution. Therefore the following proof can be advanced without placing intellectual stakes in particular evolutionary mechanisms as they are currently conceived

***

In my post on The Mathematics of the Spongeam I used the following equation as a way of talking about evolution:

This is basically a diffusion equation with an added term and where Y represents a population value subject to diffusion in a multidimensional space. The first term on the right hand side is the ordinary diffusion term expressed in n dimensions and this is a way of representing a random walk across configuration space. The second term V Y introduces the net result of multiplication and death at a particular point in configuration space. As I pointed out in my previous post on this equation, although it encapsulates many complexities it is a huge simplification of reality; for example the diffusion constants in front of the "house²" symbol could vary with dimension and coordinate. Moreover, as  I also pointed out in my previous post on this equation, as it stands it doesn't explicitly acknowledge an important potential non-linearity. Viz: that if V depends on the environment then it is clear that the value of Y is part of that environment. Therefore V is not just a function of the coordinate system but also of Y. However, in spite of these simplifications to the point where the whole equation is almost a cartoon I can nevertheless use it to express my problems with Moran's brand of thinking.

The diffusion term in the above equation expresses that important feature of evolution; namely, that it proceeds in small random steps. Given this picture one thing becomes very clear: It is fairly intuitively obvious that pure random walk would never produce what we are looking for in terms of the highly organised complexities needed for self perpetuating, self replicating structures. If in the very remote chance any organised complexity was arrived at via random walk that walk would ensure that it would very quickly dissolve back into the sea of general randomness.  For organised structures to have a chance of persisting for any length of time we need the second term on the right hand side of the equation, that is V Y. Where configurations have a net replication V will have a positive value; where populations have a net decay V will have a negative value. However, since we expect that the number of configurations which have sufficient organised complexity to return a net value of V to be very small compared to the whole of configuration space then this doesn't give us very many viable self replicating configurations to spread across configuration space. Hence, relatively speaking configuration space will be almost empty of self replicating configurations; we know this simply because a successful replicator will clearly have to be sufficiently organised and the number of organised structures in the whole of configuration space constitutes a very tiny percentage of the whole of configuration space  (See my book on Disorder and Randomness).

Now here's the rub: Since evolution must start from square one (i.e no replicators) the set of self perpetuating, self replicating configurations must be fully connected all the way from the most elementary replicators to sophisticated organisms. This requires V to be such that it forms channels in configuration space favourable to replication, along which evolutionary diffusion can diffuse. To assist the imagination in visualising this connected multidimensional set I use my picture of the spongeam. Viz:

The above is a three dimensional visualisation of what is in fact a multidimensional object; it is  system of very tenuous fibrils spanning across largely empty space. The spongeam may or may not actually exist, but we at least know this; it is a necessary pre-condition of evolution, at least evolution as it is currently understood.

This picture of the spongeam actually prompts a pertinent question: If replicators, which necessarily have to be highly organised, are so utterly dwarfed in number by the immerse size of configuration space, are there actually enough of them to populate configuration space with a class of points sufficiently connected to allow evolution by diffusion from square-one to advanced and complex organisms? I have my doubts but for the sake of the argument we will proceed as if there is a spongeam sufficiently connected to allow diffusional migration from simple structures to complex self replicating organisms.

Whether a spongeam exists to facilitate evolution depends very much on the form of V.  It is possible to "cheat" and simply patch pathways and channels into configuration space ad hoc style to ensure that a wide range of replicators can be reached by diffusion. But such special pleading doesn't seem to be how our cosmos works; its fundamental parts are more like a construction kit such as meccano where a few fairly elementary parts and their "fixing" rules are specified and we are then left with the problem of whether such a system, given the diffusion dynamic, can build or "compute" general replicators. Let me say again: I have my doubts, but that's really another story**. Suffice to show here that given current understandings of evolution Larry Moran's assertion about it being a process of "pure randomness" is entirely misleading.

In the spongeam picture the difference between natural selection and Moran's emphasis on the randomness of the evolutionary walk is a fairly minor difference: In Moran the neutral diffusive evolutionary  "channels" of the sponeam are akin  to "level" pathways where there is no bias pushing in a particular direction. Natural selection on the other hand corresponds to the case where the pathways have a kind of slope by virtue of a changing slope in V which means that the random walk is a walk with a bias in a particular direction and therefore proceeds more rapidly in that direction. But as far as my equation is concerned this is just a variation on a theme and evolution may proceed under both circumstances. (The irony is that if evolution has occurred then I'm very favourable to Moran's concept of evolution!: It minimises the idea of evolution as a fight for survival and instead something more like the drifting apart of languages  when communities are separated)

But whatever! My the main point is that all this exposes Larry Moran's misleading view on evolution; for whether evolution is neutral or biased both scenarios require a spongeam envelope that introduces a considerable information constraint; that is, evolution is a process which in this sense is far from maximum randomness and presupposes a highly organised constraint spanning configuration space.

I'm not here arguing whether or not the spongeam actually exists in the real world: Rather I want to make the point that if it does exist - as it must if evolution has occurred as currently conceived - it does no justice whatever to describe the process of evolution as if it is pure randomness at work: Pure randomness would lead to nothing: If the spongeam does exist in our cosmos then presumably its convoluted network of fuzzy pathways is determined by the set of fundamental particle interactions.

That evolution must start with a huge information bias has been proved generally by William Dembski. Dembski has been unjustifiably abused for his efforts; a sign of the worldview stakes involved in the debate. Although I don't accept the inference that some IDists have thence drawn from Dembski that information can't be created, Dembski's result is sound. I present a back of the envelope proof of this theorem in this paper.


Relevant Link:
http://quantumnonlinearity.blogspot.com/2018/01/evolution-its-not-just-chance-says-pz.html

Footnote

** The de facto "Intelligent Design" community are quite dogmatic on this point: For them there is little or no doubt that given the cosmic physical regime evolution, as currently understood, is impossible and that "Intelligence" of some sort (they don't elucidate what sort) needs to step in somehow (they don't elucidate how) and do its stuff of filling in the creative gaps in the creation story. The irony is that what raises a question mark over their thesis is the very existence of the super-intelligence they posit. For if that intelligence is none other than an omniscient omnipotent God then who knows what such a being is capable of: For all we know there may be a set of particle interactions which imply a spongeam and that God may have chosen that set! (although as I must repeat, I am myself doubtful about the existence of the spongeam - but I might be wrong!)