Epistolution Musing №9: Popper’s Conjecture
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 that purposefully activates genetic influences and applies them to problems. Clear examples of epistolution include embryonic development, wound healing, regeneration, cancer, learning, memory, creativity, swarm intelligence, and the placebo effect.
Recap: In the last section we covered how causal suspense can provide the motivational logic that makes us curious to interact with certain parts of the world more than others. But there are four other congruent aspects of cellular knowledge that are salient as well. This week we meet the philosopher Karl Popper and discuss how he disproved the theory of intelligence now ubiquitous in the tech industry and pop media. Popper demonstrated that the process of learning logically cannot be conditioned on statistical examples, but has to be developed by a physical process that anticipates causation in a spontaneous way, in the form of an explanation.
My five observations about cellular knowledge are that it is:
1. Causal
2. Conjectural
3. Information-determining
4. Open-ended
5. Anti-entropic
This week let’s look at conjectures.
Karl Popper, the famous philosopher of science, gave lectures in an auditorium at the London School of Economics after the second World War. Dressed nattily, he stood at the podium and commanded his students to “Observe!” As the student shifted restlessly in their seats, Popper then stood silently in no particular posture, simply reading over his notes and saying nothing for a long interval. After a time the students looked around at each other, they struggled to discern what his statement meant. What should they observe? Was it something else in the room? Where should they direct their attention? It was an unnerving situation.
Popper’s point, which was also the overwhelming thrust of his philosophy, was that theories or truths cannot be built from observations. One must already possess a theory in order to know what to observe. Before Popper, the consensus was that scientists look at the facts, and see what the “preponderance of the evidence” is, and from this they build their theories. This false is still widely believed despite Popper’s advances. It makes intuitive sense, but it can’t be right. Popper spent his career refuting this idea. He noted that one cannot build a theory from facts because there are no facts without a pre-existing theory. In order to say what has happened anywhere, with anything, you need an explanation that divides the world into entities, concepts, and forces and things like that. You need all these to tell you where to look to apply rationality. A new theory comes not from facts but from problems. These are the cases where the existing theory is either in conflict with itself or with things that happen in experiments. When you have problems, you can invent a new theory to solve them with a conjecture, or guess.
Popper was trying to solve the problem of induction, a longstanding philosophical issue in rationality. Since ancient times, many philosophers had understood truth as something like “beliefs, justified by events.” In the seventeenth century, David Hume had noted that no matter how many examples of a true statement are confirmed, nothing about this ensures that they won’t be refuted in the next instance. This meant, he argued, that it was entirely irrational to justify beliefs by pointing to the cases in which they had been confirmed. The sun may have risen many times, but this was no logical proof of the next sunrise. His famous example was the whiteness of swans. No matter how many swans one saw that were white, it was no proof that swans couldn’t be another color. When the explorers reached New Zealand, they found swans whose natural plumage was jet black, overturning all these “justifications.”
This was an unsettling claim. Hume’s work really destroyed the rational basis for scientific claims. Of course it was not a problem in practice, but only in understanding. For several centuries, experiments and theories continued. Progress did not screech to a halt because of one philosophy paper. But this only shows that life, and science, can still proceed if we don’t know quite what we are doing. Understanding what we are about is still key to making better progress. All human endeavor rests on a bed of fundamental claims and assumptions, even if they are unknown or unacknowledged. The more we can elucidate these assumptions, the better we can accomplish our aims. Hume’s problem had to be overcome, and Popper brought it some resolution.
Popper’s scheme for science, (and knowledge in general) was that we must first make conjectures, and try later to refute them. Truth, for Popper, was real, but he felt we can never establish it beyond question. We can only approach it closer and closer by making better guesses. Popper felt that the demarcation between science and nonscience was that science was where it was required that your ideas be falsifiable. Scientific, falsifiable ideas could always rest on unfalsifiable assumptions (e.g. there is something rather than nothing), but science was fundamentally an activity where we try to invent new guesses, and then find out if they either contradict themselves in logical analysis, or if they can be contradicted empirically by physical experiments. The act of conjecture, for him, was prior to logic and to empirical observation.
This way of thinking is simply brilliant. Without a Popperian understanding, we are stuck confining the process of developing “facts” to the current paradigm, within which all the existing “facts” will have been observed. Without the recognition that a wild new explanation is necessary before we can assess the current facts for credibility, we are left without any means of developing any knowledge that is truly or radically new. Popper gifted us with a means of making much more rapid progress in science, if only we will now take him seriously.
Creative, bold new explanations that give us new places to look are the ticket for this Popperian sort of progress, not the exhaustive gathering of data. Data are important, but only when there are two explanations to be compared. Too much current scientific work blindly gathers data without first clearly articulating which plausible explanation one is attempting to refute by the results. This muddies the water with lots of pointless, undirected analysis and useless received ideas. Popper showed that if one refuses to clarify or state explicitly the consequences of one’s theories in order to protect them from criticism, it is impossible to prove them wrong and move on. As Einstein said, the very best result for any successful theory is for it to be replaced.
In Popper’s view, explanations were special things that must meet particular characteristics. They must be internally consistent, broadly applicable, and deeply informative. Some more recent Popperians have improved his views a bit, such as David Deutsch’s proposal that explanation should be “hard to vary” and Nicholas Maxwell’s observation that the preference for unified explanations over disunified ones is irrational, but vital to our way of thinking. I personally think of the requirements for a good scientific explanation as fitting the acronym PLITH. It should be Parsimonious, Logical, Informative, Testable, and Hard-to-Vary.
Brilliant as it is, Popper’s conception inherits an obvious weakness, a weakness that was already inherent in Platos’ view of truth that Hume was criticizing. If conjectures are the path closer to truth, what are they? How are conjectures formed? I don’t mean in the sense of how they are spoken or written, but from what fundamental process do they arise? The answer (by definition) is that they arise through the subjectivity of point of view created by epistolution. We don’t know the actual mechanism yet, but we know that biology can do it, and given what we know about biology, it seems likely that all beings can do it.
When I realized this congruency between Popperian epistemology and biological theory it was a real lightning bolt for me. I wrote an email in January of 2021 to both David Deutsch and Denis Noble urging them to get together for a long cup of tea so that they could design the research programme for epistolution. It was not a joke. In a very tangible sense, I think of organisms as conjectures. I don’t mean I think of them as making conjectures; I think of them as actual physical embodiments of conjectures. If conjectures can be held physically in brains, they could also be held in entire organisms. Just as written conjectures can be refuted by comparison with data, organisms can be corrected by collisions with the reality of their surroundings. An explanation is a map of causal suspense drawn all over the world, telling us “what wiggles what.” In exactly the same way, an organism can be thought of as a conjecture about whatever it understands. In the same way as Popperian scientists, organisms direct their activities toward the most likely areas they can find that can both provide surprise and mend it, increasing the confidence of their subjective view of the truth.
Next week we look at Claude Shannon, the founder of information theory, and we explore biology’s ability to bind information, the thing that links DNA as a code to the codes that we write in our software.
Be Kind, and Be Brave,
Love, Charlie
