Confusion and convergence, but no "myth of randomness." Chapter 2 of From Extraterrestrials to Animal Minds

The concept of randomness is caught up in evolution, in two broad ways. The first and most famous aspect is the oft-misunderstood randomness of mutation. The second aspect is the role of chance in the trajectory of evolution. It is this question—is evolution predictable, or is it a random "drunkard's walk"—that Conway Morris tackles in the second chapter of From Extraterrestrials to Animal Minds: Six Myths of Evolution. The chapter is called "The Myth of Randomness."

The chapter is a chaotic mess and ends without a clear argument, much less a convincing one. Conway Morris wants to tip the scales away from "randomness" and toward "cyclicity." From the second paragraph of the chapter (page 43):

Although through geological time increasing degrees of biological complexity and integration are undeniably the case, superimposed on this is an intriguing cyclicity: plus ça change, plus c'est la même chose. It transpires that evolutionary history is very far from random.

Rincon Mountain Wilderness

Like he did in Chapter 1, Conway Morris takes a profoundly interesting, long-standing question in science, a question that currently inspires brilliant writing and experimentation by evolutionary biologists, and paints one of the possibilities as a "myth." Like he did in Chapter 1, he erects a strawperson. Unlike in Chapter 1, the strawperson is not a laughable nonexistent entity but instead a fuzzy caricature of a major factor in evolutionary history—the combination of chance and contingency. Fortunately, unlike in Chapter 1, Conway Morris provides a poorly integrated amalgam that few laypeople will understand. Thus, even those inclined to cheer the immolation of the strawperson will find little more than stuff like this (p. 56): "The clear implication is that beneath these entirely plausible factors there are deeper organizational principles at work and of which we know very little at present." That sentence is typical of the chapter, which reaches its nadir at an invocation of "particle physics or the periodic table" as evidence that physicists embrace the notion of "a deeper order of the world" while biologists struggle to do the same. In his ardor to preach on these "deeper" things, Conway Morris obscures the grandness of the question and at times distorts what working scientists know and do. It was on finishing this chapter, a few weeks ago, that I regretted buying and reading the book.

This chapter (and perhaps the whole book) seems to lack an identifiable audience. Lay readers will not follow his reasoning and will not understand why his examples are even relevant to a "myth of randomness." On the other hand, those who know the literature on this fascinating topic will already know that randomness is not a "myth" in any sense of that word, and will already know that we have excellent evidence for both "sides" of the question. Inconveniently for ideologues and religious apologists, this is not a debate that a doctrinal faction gets to win. Randomness is a major factor but convergence shows that adaptation and principles of design are deeply involved in evolutionary trajectories. Reasonably well-informed people already know this.

So, is there anything new or interesting in the chapter? "Cyclicity," what Conway Morris calls "the eternal return," is a bit more than mere convergence. It is repeated convergence, in which evolution returns to the same solution over and over. A classic example is C4 photosynthesis, which has been invented/discovered more than 60 times. This is an interesting subject, but Conway Morris rambles through the topic with no apparent question or purpose beyond asserting a "clear implication" that there are "deeper organizational principles at work." He seems impressed by the fact that there are evolutionary "hotspots" in the tree of life, and points briefly and vaguely to "Vermeij's observation that some niches appear to be potentially available but remain unoccupied" while avoiding what Vermeij thinks this might mean. The uneven occupation of Design Space is indeed interesting, but it's not a new subject for Conway Morris or anyone else.

Okay, but does Conway Morris have interesting takes on randomness? Sadly no: he is mistaken about the meaning of work on genetic hotspots of phenotypic variation; he writes that "in at least some cases mutations are very far from random but rather tend to be concentrated in so-called hot spots." This is a garbled misstatement, both of what "random" means in the context of mutation and of what the cited work is actually about. Indeed "randomness," evidently the topic of the chapter, is a term Conway Morris at times seems not to understand. For example, from different work than I just cited, we know that mutations are not equally distributed in genomes—indeed it has long been known that there are particular mutations that happen far more often than "normal" and many of these are well understood (keyword: breakpoints). This says nothing about whether mutations are random, because the term was never meant to claim that all mutations in a genome are equally likely. That quote about "concentrated in so-called hotspots" suggests that Conway Morris doesn't know what random mutation means.

Maybe that doesn't matter, because this chapter is really not about randomness. A clue to the source of confusion can be found in the strange introduction on page 43, wherein the author takes us to a casino where some countess is playing the Game of Life. She loses ("a gasp of surprise") then "walks into the darkness of extinction." (Those are high stakes!) This happens next:

...hitherto shadowy figures now join the game. But they are very far from being unexpected arrivistes—rather, they were not only waiting in the wings but turn out to be yet more accomplished players. The rules remain the same but at a yet faster pace.

This is really not relevant to randomness. But it helps us understand the author's project, because the casino depicts the theme of the next chapter (the "creative" role of mass extinctions). What Conway Morris wants us to imagine is that the Game of Life is somehow rigged—that success involves following established pathways through Design Space, pathways so vanishingly rare that they render nearly all of Design Space uninhabitable. And he wants us to believe this is true in organism Design Space and in molecule (protein) Design Space. I see two obvious problems with his reasoning:

• Rarity of "good design" is both relativistic (what is "rare"?) and expected.
• Neither contingency nor chance is relevant to the question of the "density" of Design Space. Those are concepts about how evolution explores Design Space, whether that space is overflowing with viable options or resembles a trail map of an Arizona wilderness.

Once we identify the author's misconceptions about randomness, the chapter resolves into these themes:

• a nice exploration of some examples of convergence, which surely suggest tracks through Design Space but are simply the latest writings by Conway Morris on a now well-established topic; and
• a flawed attempt to establish evolution, including molecular evolution, as a journey on a "knife-edge of biological existence, the narrowest of tracks that thread their way across an otherwise desolate landscape of nonviability."

With the caveat that it's not clear how to dispute terms like "knife-edge" and "narrowest of tracks," I think we can now say that Conway Morris is wrong about that second theme in the realm of protein evolution. It was dubious when he wrote it*, but as of May 2022 it became nearly indefensible. Molecular evolution, it turns out, is easy. I'm writing a book about that, and I'll start writing about it here, soon.

So finally, is there some orthodoxy that Conway Morris is challenging in this chapter? No. He's staking out a position that seems extreme, and that is flimsy in important ways, in a subfield of evolutionary biology that already knows that both extremes are wrong. It looks bad and is reason enough for you to avoid reading the book. But is there some interesting biology in the chapter? Yes, of course, and my favorite stuff was about convergence and innovation in the proteins that carry and store oxygen in animal bodies. The section on the machinery of photosynthesis was chaotic and, like everything else in this chapter, an expansion of the author's previous writings on exactly the same topics. But who doesn't love photosynthesis?

Next up is the first chapter that was worth reading, about mass exctinctions. As I'm sure you could guess by now, there's no "myth" exposed in the chapter, but finally Conway Morris succeeds in making a case for an adjustment in our thinking.

---

*He doesn't say when he was writing but there are citations to 2021 papers. The book was published in mid-2022.

Image credit: Rincon Mountain district trail map. From Wikipedia.