I think it’s time that we deal with the “random chance” creationist canards and related garbage once and for all. And in that vein, I’ll present here the draft of a forthcoming detailed exposition on the subject.
It’s blatantly obvious that none of the usual suspects either understand what the word “random” means in rigorous scientific circles, or have any inclination to learn this, not least because doing so would destroy their worthless apologetic lies at a stroke.
In rigorous scientific circles, the word random has a very specific meaning. Namely, a random process is any process whose outcome conforms to a well-defined probability distribution, and which can be modelled by a Markov chain process.
This will, of course, need some explanation for those unfamiliar with the above definition and its import.
First of all, a Markov chain process is any process that involves well-defined transitions between states, in such a manner that there may exist more than one possible transition between any two arbitrarily chosen states. As a corollary, in the absence of any information allowing us to determine which of the multiple transitions actually occurs between those states in a given instance, the multiple options are assigned probabilities. I shall now link to a useful and more detailed exposition that can be read at leisure.
There are two essential points to make here. The first being that in the case of a Markov chain process, each of the transitions is frequently considered to be a deterministic process leading to the state change. For example, if we examine in detail the chemical reactions leading to a particular type of point mutation in DNA, we may find that there are three sets of chemical reactions leading to that state change. Each of those chemical reactions is a deterministic process, but if we do not have information telling us which of those three options took place in a given instance, then we have to model the process as a Markov chain process. While the different options have probabilities associated with them (which in this case, can be determined by an appropriate experimental setup), each option in this example is a deterministic process - the very opposite of the duplicitous “happened by chance for no reason” creationist caricature.
The second point is that since the outcome conforms to a well-defined probability distribution, this means that there must exist a constraint upon the outcome, and where there exists a constraint upon the outcome of physical interactions, there is a mechanism underpinning this. We may not know the details of that mechanism, but we can be sure it exists, because the only other option is magic. The moment we introduce magic into the discussion, we’re not dealing with science any more, or indeed, any proper systematic understanding of our surroundings.
Indeed, that principle above, that a mechanism must exist governing the outcome in some way, frequently leads to fruitful searches for said mechanism. Which frequently takes one of three forms - a low pass filter, a band pass filter, or a high pass filter.
Well, guess what? Natural selection (or indeed selection by human agents) is a high pass filter, with respect to relevant criteria. The criteria in each instance may be different, but the principle remains the same. Namely, that some criteria are deemed, either by nature or by human agents, to be measures of fitness, and selection propels the entities being selected up the slope to a local fitness maximum. Indeed, one of the great insights in biology, has been that selection can arise as a result of impersonal, non-sentient natural forces. The presence of pathogens, predators, even weather, can affect the outcome with respect to the growth or otherwise of a population of living organisms, and the observational data informing us of this is now voluminous.
The above is routinely pressed into service in simulations, using what is known as the Monte Carlo method, a topic that has entire textbooks devoted to it for those who genuinely wish to learn more.Those simulations frequently produce outcomes that are as far removed from “happened by chance for no reason” as it’s possible to be, outside of actual deterministic physics.
Indeed, in many areas of science, the word random, given that above definition, is used as a shorthand for “we don’t have the audit trail of data telling us what happened precisely in these instances, so we’ll use a Markov chain model to assist our understanding thereof”. In areas where we don’t have a full audit trail of data, Markov chain modelling becomes in some instances a necessity, and its no less valid than any other scientific tool - indeed, the list of applications is now huge, and this link provides a brief flavour of the applicability of Markov chain processes.
There’s a large body of rigorous mathematics underpinning the use of Markov chains, and indeed that link immediately above provides some examples of that underpinning mathematics.
So, can we, once and for all, take the duplicitous creationist caricature of “random” out into the courtyard and shoot it? Ideally along with all the duplicitous caricatures they peddle?
