Randomness and Functional change

Defining what is meant by ‘random’ is itself a major field of enquiry in mathematics, computation and in science generally. The question whether there are truly random events in the universe is a vexed one, lying at the heart of theories of quantum mechanics. Probably, we will never know, perhaps cannot know in principle, the answer to that kind of question.  But it warns us that defining randomness is not easy. I also say that in the lectures.

The best way to sidestep the deeper problems is to ask the question ‘random with respect to what?’ In evolutionary theory that makes the problem much simpler. Both Neo-darwinists and their opponents can then agree that what is really meant is ‘random with respect to physiological (phenotypic) function’. That is so because one of the central tenets of Neo-darwinism is the exclusion of any form of Lamarckism, the idea that function, or functional improvement, can influence inheritance, i.e. the inheritance of acquired characteristics. By contrast, a Lamarckian must maintain that at least some changes are not completely random with respect to function.

In the article, and inevitably in much shorter form in the lectures, I approach this question in three stages. The first stage is to establish that genomic change is not random with respect to location in the genome. The reason for asking that question first is that without establishing that there are preferred locations of change, the argument for any kind of functionally-relevant change cannot even get off the ground. The only way in which such a change can occur is through influencing the physical and chemical properties of the genetic material. Preferred locations of change are therefore a pre-condition for functional change to be possible. If all locations in the genome were equipotent for changes there would be no possibility for functionally-relevant change.

Of course, demonstrating the existence of hotspots and other ways in which change is not randomly distributed with respect to location does not, in and of itself, demonstrate any form of functionally-relevant change. The existence of hotspots could be simply a consequence of the physico-chemical properties of the genome and its associated proteins even if no functionally-relevant changes occur. Further experimentation is required to demonstrate that. I also make that clear in the article and lectures.

The second stage in the argument is to note that well-documented examples of functionally-relevant genomic change already exist. The best-investigated case is the evolution of lymphocytes. In response to antigen activation the relevant part of the genome undergoes very rapid proliferation. Targeted speeding-up of change is therefore one mechanism by which functional change can occur. That is true even if the individual changes at that location are random. The functionality lies in the targeting of the location. A similar targeting occurs in P element homing. And targeted reorganisation of the genome by speeding up change in selected locations occurs in bacteria subject to starvation. All these mechanisms are referred to in the item Relevance to physiology.

The third stage in the argument is experimental demonstration that the inheritance of acquired characteristics occurs. There are now many examples of that. See Transgenerational Inheritance.  Those experimental results require functional inherited change, either genetically or epigenetically, even if we do not yet know the molecular mechanisms. All these stages of the argument are necessary. Some of the critics have mistaken the first step for the whole argument!