In two recent posts here and here I tried to show, via the route of Thomas Aquinas’s Fifth Way, that “randomness” as it is actually found in the world is a sign of order, not disorder, and therefore points to God no less than does final causation in nature.
Inherent in my argument was that, in any scientific context, to regard “chance” as some uncontrolled force is completely erroneous. Instead, one needs to see it as a probability distribution (whether known or as yet unquantified) defined by the particular system of causes which is under consideration.
And so in my “index” example of a coin toss, it makes no sense to talk about chance without a context. Given the universe as a whole it’s extremely unlikely that one will find a coin of any sort, and only half as unlikely that it will be showing heads rather than tails. But in the context of a designed system, and under the circumstances of a deliberate human choice to operate it, things are different. As I said in the second recent post linked, each event in a coin toss is fully determined by the initial conditions and physical laws, but the system design introduces the means to obtain a near-perfect 50:50 probability distribution.
At this stage I want to consider this not as apologetics, arguing like Aquinas for the existence of God, but theologically, assuming that God is there as “First Cause”, in order to question the use by some theistic evolutionists of the idea that God “uses” “random and blind” chance in his creative work – especially, of course, in evolution. This is prompted by being reminded of arguments, such as that by Kathryn Applegate (and hammered by the infamous John/Melanogaster) at BioLogos back in 2010, that the human immune system depends on just such “random and blind” chance, which assertion may then be generalised to support standard Neodarwinian evolution by “unguided” chance.
Kathryn was also the author of a similar, much simpler, suggestion about God’s use of Brownian motion in the assembly of viruses as an case of God’s use of “chance”. I’ve pointed out before the triviality of this last example, but one can illustrate it by another of the examples from my recent post, the statistical behaviour of gases. On Kathryn’s reasoning, God would “make use of randomness” in the movement of molecules not only for virus assembly, but for the sun’s operation, the wind in the trees, all that breathes air and everything, in short, that involves gases above the molecular level. It may be true, if one ignores the deterministic movement of individual molecules, but it’s pretty misleading therefore to put all the physics of gases down to “blind chance”. It’s actually the harnessing of a probability distribution within a law-driven system.
Returning to the immune system, when I first came across this argument, my own half-remembered medical training on immunity, received in Cambridge just a few years after the basic mechanisms were unravelled, made me quite certain that this is a very poor way of looking at things. The immune system clearly uses extremely constrained, and bafflingly sophisticated, means to produce effective responses to the widest possible range of pathogens, and those means include highly targeted hypermutations as the second stage “tuning” response to pathogens detected by the binding to them of already near-universally diverse B and T cells.
I discovered a long response to Applegate’s treatment of the immune system from immunologist Donald L Ewert (a former associate editor of Developmental and Comparative Immunology), but considering the system in the way in which I did for the “coin toss randomizer”, we can make a direct comparison using those terms.
The immune system is teleological, in that it has the clear function (or “end” if one wants to be more Aristotelian) of dealing with a multitude of immune threats. All the events involved in it are subject, as far as is known, to deterministic natural law, but the system as a whole incorporates a specific probability distribution at key stages to ensure, to a pretty high degree of certainty, that all possible immunological bases (a large but finite number) are covered. If the probabilities were not tailored to the system, chance in the sense of “blind randomness” would not be at all likely give a complete distribution of the needed specificities of basic immune cells (whose absolute numbers are of the same order as their variety), and nor would the fine-tuning mechanisms produce sufficient useful, rather than counter-productive or even harmful, variations. Disorder, remember, cannot be represented mathematically, whereas probability can because it is a form of order.
To put it another way, the system’s “knowledge” and utilization of the probability distribution of its gene mutations means that the randomness is as constrained as it is in a coin toss, and the teleology of the system removes any idea of “blindness”, in the same way that a referee knows and harnesses the coin toss’s 50% probability when he uses it as a randomizer to start a football match. One can only justifiably use the word “blind” in the sense of denying conscious intelligence to the system, as one would anyway even if it were 100% determined, such as the function of the heart, say, or the teleology of a strong chemical reaction.
As I said before, probability distributions are functions of teleological systems. Indeed, they can only be identified or quantified in relation to such systems. How likely is it that life would form on earth by entirely natural causes? It depends entirely on how God designed the system. If there is no inherent tendency of the constituents of the earth to form life, then the probability is zero, and either God is a fool or intends to create life in some non-natural way. If such a tendency is present, and the quantities of the right materials and time available is right, then the probability is 100%, and any talk of “blind” or “random” forces is meaningless. The setup of the system causes the result with certainty – it’s a straightforward act of creation through secondary causes.
But if, assuming we had sufficient information to calculate it, the probability of the earth-system forming life were some other figure than 0% or 100% then, as an act of creation, something odd would be going on. Perhaps God is a gambler. After all, we know why a referee wants a 50:50 probability on a coin toss. We know why the immune system needs the particularly wide variations in its immune cells (which is pretty much identical, incidentally, across the diverse immune systems of the animal kingdom). Why would God, though, wish to play percentages in creation? The question is academic, though, because nobody has any way to calculate a probability distribution – which makes any assertion of “chance” in connection with the origin of life meaningless metaphysical (and ideological) speculation.
The use of “blind chance” in connection with evolution is therefore, in the same way, meaningless unless backed with a probability distribution. It’s valid, in my book, to attempt to estimate such things provisionally in connection with proposed mechanisms for particular processes where the “working” can be shown. So for Eugene Koonin to reason that DNA replication is essential before natural selection can start, and then to calculate that it is beyond the resources of one single universe, gives a basis for agreement or refutation. In the same way, if evidence is presented that the frequency of benefical mutations in cells is x, and that the evidence is that they can accumulate stepwise y% of the time, one knows (or can dispute, from the evidence) if the system we call “life” will evolve long-term. But if it will, all you have learned is that the design of the system was such as to make it work on that probability distribution.
What appears to be quite invalid is the assumption that, on some self-evident philosophical principle, random variation and natural selection could achieve any result whatsoever simply because it’s instinctively plausible. Some empirical crunching of probabilities is required to define your process, in order to evaluate it. If there are no limits to what is possible, the process being invoked is supernatural, rather than natural, because all natural processes have limitations. And those limitations can’t be determined simply by assuming that every feature of life ever found must be a result of what you’re seeking to establish. Common though that habit is.
In fact, the most interesting thing I gained from Ewert’s essay was that it is possible to speculate on the parallel between the immune system’s operation and evolution more widely, but not in the way that Kathryn Applegate supposed in saying that, since blind and random changes make the immune system function, then blind and random forces could be what God uses to create through evolution. Ewert points out that in fact what systems biology shows is that the immune system is a sophisticated and highly targeted process that harnesses a particular probability distribution towards a highly specific end (that is, its protective function).
If one wanted to draw a parallel with evolution by mutation and natural selection, then, one would need to draw it as a highly sophisticated and targeted system that harnesses a specific probability distribution with particular outcomes in view. Evolution would then be, as James Shapiro and others have suggested, a “physiological” mechanism for facilitating targeted adaptation, and by that same token guided and purposeful, by internal or external teleology, and not undirected and blind.
Now that, for some, is more likely than to result in ditching the use of the immune system as an illustration of randomness in evolution than in admitting teleology. Which would at least be intellectually honest. But for a theistic evolutionist, such a teleological system seems a whole lot more consistent with traditional theistic creation than with talk of “blind” or “random” events.