Epistolution Musing №4: Epistolution Appears
Dear Friends,
This letter is part of a weekly series of brief thoughts I would like to share with you, either because I’ve come across your related work in biology or because you’re a person I like. I discovered an interesting problem in 2019, a problem I can’t forget. Epistolution is the unknown biological mechanism necessary to explain purposive activity that can’t be accounted for by genetic influences.
Recap: Last week we noticed that organisms cannot be making decisions based purely on natural selection because this problem is outside of their domain. Before delving further into what this means, let’s focus a little more specifically on why there must be a process prior to selection that motivates life. This week we learn why selection presupposes, rather than explains, epistolution. In so doing hopefully you will begin to see a new gestalt of the world in which epistolution will no longer try to disappear from your mind’s eye.
Social scientists are accustomed to dividing influences on organisms into “genes” and “environment.” Of course, the difference between twins is environmental, not genetic. But this way of speaking is hiding a presupposition that needs to be exposed. There is something about an organism that allows a novel environment to make it seek after novel goals which are compatible with life. That’s the problem. What controls the effect the environment has on the organism such that the organism adapts to it, rather than randomly malfunctioning? Why do organisms produce adaptations in novel conditions rather than going haywire?
Some extreme selectionists, like the Neo-Darwinist Richard Dawkins, suppose that there is no alternative to natural selection for producing adaptations, that any other mechanism is a religious claim. But this is not true. There could be other natural mechanisms that could produce adaptations. We can’t know in advance that there isn’t another, and there are some very solid reasons to insist that there must be at least one. It’s not an act of blind faith to conjecture that science hasn’t yet discovered everything it will discover.
From the selectionist point of view, there might be, in principle, no general mechanism for learning. From this point of view learning is part of a long, slow, incremental adaptive process shaped by local contingencies. From this sort of process the details may not be generalized into rules. If natural selection is the only nonrandom force shaping life, then you probably can’t really know “how” learning happens. Learning is just an outgrowth of the parochial cellular knowledge that evolution has accumulated in the heritable material through arbitrary accidents over eons and eons. From this point of view there is no such thing as epistolution.
What is wrong with the idea that evolution “found” learning? This commonsense view has predominated since the genetic era began, and most biologists see no reason to discard it.
First, because evolution by natural selection is blind. Learning is, by definition, adapting to conditions you have never encountered before. Some evolutionists try to defray the force of this argument by pointing out how the conditions organisms can learn about might resemble in some way conditions that may possibly have happened before during the long history of the lineage. This dodge only gets you so far. Pre-adaptation, for this is what this idea is called, can only account for a very partial sort of learning, one that would get weaker over time rather than stronger. Instead, in organisms we see learning that builds on itself recursively. Practice makes perfect. Blind selection doesn’t care whether an organism persists or not. It’s inherently subtractive. It operates by eliminating failures, not by seeking out successes before they arrive. The hard question is not why some lineages survive better than others, it’s why all lineages haven’t evolved extinction.
Second, because of stochasticity in the cell. Biology is a process that happens in water. Remember we are looking for something that will account for incredibly detailed learned changes in an unbelievably complex organism, changes that make it adapt in myriad ways to changing conditions. Specifically, we are looking for the ways that this process is altered differently in the case of two genetic clones like Anna and Amanda. In order for these minutely detailed changes to have been scripted by the genome, they would have to be encoded in something that worked like the many logic gates of a microprocessor, and this just can’t be true about a watery, stochastic cell. Adjacent cells express specific proteins at wildly different rates.
The Turing principle tells us that “an organism is a computer” in the sense that in principle, if you had a universal computer with enough processing speed and memory, its living process could be rendered down to an arbitrary level of detail. This means that we can, and will, build artificial epistevolvers (more about this later posts.) But the Turing principle doesn’t tell us what that living process is or how it works, it just expands the definition of “computer” far beyond any practical meaning. In order for the genome to have precise control over the organism, for it to contain that many logic gates, biology would need to be completely inflexible at a low level, exactly as today’s microprocessors are. There is no plasticity in a hard drive, every part of the system must be precisely fixed at each of billions of logic gates for the program it holds to compile.
Third, because something very much like epistolution is required for natural selection to begin in the first place. In order for individuals with heritable variations to show differential reproduction, individuals have to exist. An individual cannot simply be presupposed, it has to be explained. Forming an individual self requires a self-contained process that adapts to its environment to some extent. The self has to remain distinct from its surroundings, it has to utilize energy to reproduce, it has to defend its internal organization from perturbations, and it has to manipulate heritable particles in a way that sustains the integrity of their influence across generations. This sounds a lot like epistolution to me.
Natural selection doesn’t just keep popping up everywhere in the universe. Epistolution provides the basic conditions from which a search for function could begin. DNA alone in a petri dish, as Denis Noble likes to remind us, does not spring to life. For anything to activate and utilize heritable particles such that they get inherited, it must already have been an epistevolver. When we discover the epistolution mechanism, we will probably find that it can theoretically turn many different things into genes, but using the most widely shared libraries of genetic tools gives an advantage and that’s why DNA has become universal.
This would mean that the key to the “origin of life” problem isn’t in ancient biochemistry; it’s in all biochemistry. The question isn’t what were ancient early cells doing that caused them to replicate genes; it’s this: what are all living cells doing in common today that makes genes useful?
The theory of life contains a gap either way you have it. If you accept selectionism then the gap is this: how are the arbitrary details of evolved life arranged to accidentally produce the process resembling learning that makes phenotypes from genotypes? If you accept universal epistolution the gap is this: what are the principles of the universal learning mechanism that has enabled all life? In either case, there is a tremendous mystery to solve. The main difference is that if you accept the existence of a universal epistolution mechanism you will possess a much stronger scientific explanation for the existence of life.
Next week we look at Conrad Waddington’s research, and how it reveals a remarkable revolutionary possibility that makes the theory of evolution more plausible.
Be Kind, and Be Brave,
Love, Charlie Sent 12/27/23 edited 1/11/24
