Philosophical Dead Ends

John Dupré reviews Richard Dawkins’s “The Genetic Book of the Dead” and Sara Imari Walker’s “Life as No One Knows It.”

By John DupréNovember 30, 2024

• Philosophy & Religion

WHAT MAKES A successful popular science book? One could not find a better paradigm than Richard Dawkins’s 1976 best-selling classic The Selfish Gene. This book presented a quite widely if not universally accepted interpretation of the so-called “Modern Synthesis,” the set of ideas that have constituted much evolutionary thinking since the mid-20th century. Because the theoretical science that drove the Modern Synthesis was population genetics, the mathematical study of the evolutionary trajectories of genes, it was natural to think of evolution as a theory about genes, and Dawkins took this view to a somewhat extreme conclusion, painting evolution as genes striving to replicate themselves—a.k.a. the gene’s-eye perspective. Individual genes were seen as competing with one another to have the most positive effects on the organisms in which they resided, and thus to become more prevalent in the next generation.

This was an intriguing vision, and Dawkins presented and developed it in elegant and engaging prose, showing an exceptional talent for providing illuminating metaphors and vivid examples. The outcome was a spectacular success: over 48 years and a million copies later, Amazon tells me that it is currently the number three bestseller on the topic of evolution.

As Dawkins has often told us, a well-adapted idea will tend to be selected and become more prevalent. The Selfish Gene was as well-adapted as a book can be, and it has illustrated this principle, though by deliberate print reproduction rather than biological reproduction. In the subsequent half century, Dawkins has written a number of books that pretty much tell the same story as his first, and they have also been met with great success. (I should perhaps give special mention to his second book, 1982’s The Extended Phenotype, which extends the basic theory in important and interesting ways, and is generally the work most admired by professional evolutionists.)

Dawkins’s most recent book, this summer’s The Genetic Book of the Dead: A Darwinian Reverie, is the latest iteration of these ideas. Since individual books are already indefinitely reproducible, a new version requires a new hook—a novel mutation?—and in this case, he provides it through the idea that an organism’s genome is a kind of palimpsest, layers of which can be peeled away to reveal the history of the organism’s lineage through the adaptive challenges that the genes evolved to address. The gene’s-eye view, to put it another way, is backward-looking. Though this is a pleasant enough conceit, it is in the end a fairly trivial permutation of the original vision from The Selfish Gene.

There are, of course, many new examples, and the illustrations by artist Jana Lenzová are a delightful enhancement. The book could serve well as an introduction to some basic evolutionary ideas such as the difference between homology (similarity grounded in common ancestry) and analogy (similarity arising from similar adaptive problems). Each of these gets a chapter with many entertaining examples. There are also chapters explicitly reiterating the main messages of The Selfish Gene and The Extended Phenotype; the book is appropriately palimpsest-like.

The major problem is that the adaptive landscape has changed substantially since the 1970s, and Dawkins has failed to evolve. Or perhaps more disturbingly, and anticipating some similar reservations about the other new book reviewed here, Sara Imari Walker’s Life as No One Knows It: The Physics of Life’s Emergence, the niche for a popular science book is mainly to provide a good story, even if it is a bit simplistic and occasionally economical with the science.

¤

Dawkins is, of course, aware that his ideas have come under criticism, but he shows very little understanding of—or, indeed, interest in—the general background out of which this criticism arises. I highlight two major areas in which biology has moved past the image that Dawkins continues to uphold.

First, Dawkins remains, loosely speaking, a genetic determinist. I say “loosely speaking” because, of course, he does not believe that genes are sufficient to produce an organism. DNA alone in a test tube does not somehow turn into an elephant or an orchid. But Dawkins does take genes to be what matters. The additional necessary resources—oxygen to breathe, parental care, and so on—are background conditions normally sufficient to allow the genes to do their causal work. A crucial consequence of this is that development doesn’t matter for evolution. If the phenotype is fully inscribed in the genes, then it makes sense to think of evolution as ultimately a sequence of genomes competing with one another. The phenotypes are proxies for the genomes that determine them.

But life is not like that. Development is complex and multifactorial. Various nongenetic factors—e.g., cultural or epigenetic (non-sequence-based features of genomes), which may have their own distinct evolutionary trajectories—play a role. This is recognized in, for example, niche construction theory, based on the realization that organisms do not evolve passively in response to the environment but actively shape the environment in which they evolve.

Dawkins does gesture toward addressing the criticism of his assumption of the causal supremacy of the gene but offers little more. He picks up a short quote from Denis Noble’s Dance to the Tune of Life: Biological Relativity (2017): “This book will show you that there are no genes ‘for’ anything. Living organisms have functions which use genes to make the molecules they need. Genes are used. They are not active causes.” He then spends 15 pages criticizing the last sentence, interpreting it as claiming that genes have no causal influence on the phenotype.

There is a great deal to unpack here, more than I can attempt in this review. The first sentence quoted from Noble’s book references an endless discussion, often frustratingly semantic. Dawkins has an inoffensive definition of genes “for” traits in terms of an increase in probability of a trait, however infinitesimal. But the expression is naturally interpreted in a much stronger way, certainly by most nonexpert readers, possibly even sometimes by Dawkins himself. The crucial point of the quote is that organisms themselves have a—or the—central role in evolution; they are not mere puppets controlled by their genomes. The causal importance of the organism itself has, I think, been amply demonstrated from multiple directions, some just mentioned. Dawkins is more inclined to rant than to engage with these findings: “It is wrong, utterly, blindingly, flat-footedly, downright wrong, to deny [the gene’s] fundamental role as a cause in the evolutionary process.”

Philosophers have struggled mightily for centuries to decide what makes some features of a complex causal nexus fundamental—i.e., the cause rather than a cause. I’m not sure exactly what makes a cause active, and this word probably provides an invitation to interpret Noble’s quote out of context. But the facts are not really obscure: in the evolutionary process, genes and organisms and much else play important causal roles. Dawkins’s attempt to distinguish genes as having a unique and overriding causal importance was provocative and interesting in 1976; now it is, I hope, accelerating toward extinction.

The failure to understand this causal complexity also leads Dawkins to dismiss a perspective almost universally embraced by philosophers of biology interested in biological causation as well as by a growing number of theoretical biologists: multilevel selection. Richard Lewontin, among others, made clear in the 1960s the necessity of this perspective. Meiotic drive is a process in which certain genes actively promote their evolutionary success by being unequally represented in gametes, or sex cells; these are truly selfish genes. But as Lewontin and Leslie C. Dunn showed, to understand the evolutionary forces acting on such a gene, one must look at the conjunction of the effect of meiotic drive and selection of organisms, with a resultant evolutionary upshot that depends on both. Meiotic drive will rapidly sweep a gene to fixation, at which point it has no other genes to compete with. So where meiotic drive survives, it is usually because the gene has a very negative effect on the organism, and it is these two effects that may create a stable balance.

Even more interestingly, Lewontin and Dunn showed that the evolutionary dynamics of the meiotic drive gene in mice, two copies of which caused male sterility (the very negative effect), did not properly predict the frequency of the gene. Rather, the solution they offered concerned a difference in the frequency of the gene in different groups of mice, allowing selection at the group level. This example, therefore, requires three levels of selection to properly understand the empirical phenomena. Group selection has remained controversial, but philosopher Elliott Sober and evolutionary biologist David Sloan Wilson, in their 1998 book Unto Others: The Evolution and Psychology of Unselfish Behavior, showed definitively that there is no objection of principle to such a phenomenon, and it seems increasingly likely that it is actually important.

Dawkins’s book fails to adapt properly to the current state of the science in another important way. On page one, he says he will save tedious repetition by treating all living creatures as honorary animals. It would be nearer to the truth to say that he takes animals as proxies for all living creatures. Though animals are of unsurprising interest to us, they encompass a small fraction of biological diversity. If evolution were a general set of laws that applied in just the same way to every evolving population, then it presumably wouldn’t matter which examples we chose to use. That, I think, is what Dawkins believes, and it points to a feature of much popular science to which I shall return at the end of this review.

But the differences between different kinds of organisms do matter. In the last few decades, there has been a great deal of interest in microbial symbionts of multicellular organisms, a topic Dawkins takes up in his final chapter. Starting from his gene’s-eye perspective, he proposes that a rigid distinction should be drawn between associated microbes that are inherited with their host (vertically) and those that are not (horizontally). The former share the fate of their hosts and are entirely friendly, while the latter have no interest in the well-being of their host and often are lethally harmful.

Plausible perhaps, but Dawkins appears unaware of the extensive discussion about the complex symbioses between almost all multicellular organisms and microbes both cellular and viral. One thing that has emerged from this discussion is that there is no sharp distinction between good vertical symbionts and bad horizontal symbionts. The status of a symbiotic bacterium or virus typically depends on the context, including where it is in the organism and who else is there. Virulence, in particular, is not an intrinsic feature of an organism, but a contextually variable property, as Pierre-Olivier Méthot and Samuel Alizon have shown. This should not be surprising if one is willing to take seriously the complex whole that is the organism; it is inevitably invisible to one who is limited to the simplistic, atomistic gene’s-eye view.

¤

There is at least one parallel between Dawkins’s perspective on life and the generally very different one offered by Sara Imari Walker in Life as No One Knows It: both books rightly emphasize the essential temporal depth of life. For Dawkins, this is the history written in the genomic palimpsest; for Walker, it is the particular trajectory, or lineage, of structures that led to this particular one out of the vast number of possible complex arrangements of matter. I will come to a rather different parallel shortly.

Like Dawkins, Walker also offers a surprise on page one. She quotes a colleague who proposes that life does not exist and remarks that she probably should have agreed with him (though she didn’t). Why would any living being make such an absurd proposal? The answer, more or less, which traces back to the pre-Socratic philosophers, is that all that really exists are the particles of microphysics (for the Greeks, atoms). The passing structures temporarily constituted by these particles are mere illusion. Only if life required some unique kind of stuff or force would it be real, and it doesn’t.

This kind of reductionism—everything depends entirely on its constituent parts, and these are all that ultimately exist—used to be common among philosophers, but perhaps remains so among physicists. One might argue that Dawkins’s obsession with genes is a relative of this view. Walker, at any rate, is right to reject the view, and the book is an attempt to explain what makes life sufficiently distinctive to be allowed to exist. Her general story goes like this: Life, first of all, is in one sense very improbable. The number of possible structures that could be assembled from atoms is unimaginably vast. Almost every possible structure has failed to exist. How did the structures that do exist come to be so lucky? David Hume proposes, in his wonderful Dialogues Concerning Natural Religion (1779), that in infinite time, every possible structure would come into existence, and the stable ones ultimately would have the chance to survive. But now that we don’t believe in infinite time, and have only a few billion years to work with, this argument won’t get off the ground. Even the macromolecules that compose living systems are enormously complex, let alone the organisms. Walker writes: “[W]hat we call life is the mechanism for making specific things possible when the possibility space is too large for the universe to ever explore all of it.” The problem posed by life, at least to those bold enough to think that it exists at all, is why.

To answer this question, Walker proposes “assembly theory,” which provides an account of the complexity of an object. The complexity of an object is the number of novel steps that it takes to reach the object in question. Objects that have already been made can be used again. The smallest number of steps that could produce an object is its assembly index. The larger the assembly index, the less probable that an object will have come into being merely by the fortuitous coming-together of its parts. Particularly significant is the case in which we find multiple similar complex objects. If there are a large number of objects with a high assembly index, this probability is effectively zero. Life, for Walker, is whatever process explains the existence of large numbers of mice or MacBooks.

This all seems harmless enough, if rather unsurprising. The frustration is that it tells us almost nothing substantive about life. In part, this is intentional: Walker is interested in a very general account of life, partly because she is interested in the possibly very different phenomena that would count as life in a wholly alien context. Life in the sense that she defines it may be nothing at all like the life with which we are familiar. Provocatively, she suggests that alien life is more likely to be found in an assembly theorist’s laboratory than on a distant planet.

What exactly will be found in either location remains mysterious. What assembly theory tells us is that the existence of multiple complex objects points to a chain, a lineage, of simpler objects that carried the capacity to construct the later objects in the chain. How complex must an object be to require such ahistorical lineage? It turns out that the answer is that it must have an assembly index greater than 15. (At this point, the number 42 will inevitably come into the minds of readers of Douglas Adams.) Though there is some suggestion that this complexity threshold provides an answer to the question of what life is, it becomes clear by the end that we still don’t know, though “much work is converging on the idea that information must be at its core.”

¤

There is a good deal to agree with and welcome in Walker’s book. The questions about how complex objects exist and why the ones that exist do, among the countless possible ones that don’t, are good questions. The attempt to provide a metric for complexity grounded in the obstacles to assembly is thought-provoking. And the emphasis on lineages rather than individual things is an important starting point for considering the origins of living systems. But from the point of view of a philosopher, there is also much to find exasperating.

And exasperation begins on page one, with the silly suggestion about the nonexistence of life. Another example is the appeal to natural kinds and the alleged exhaustive dichotomy between natural kinds—precisely delineated categories uniquely determined by nature—and subjective categories, which might have been made a good deal more sophisticated with some attention to the extensive philosophical literature on natural kinds over the last half century. The exasperation comes to a head one page from the end with the suggestion that “[u]nderstanding is muddled across the seemingly disparate concepts we refer to as ‘matter,’ ‘information,’ ‘causation,’ ‘computation,’ ‘complexity,’ and ‘life.’ Assembly theory is an attempt to see all of these as the same thing.” This strikes me, to put it politely, as complete nonsense. Many lifetimes have been devoted to trying to understand each of these concepts, and I cannot think of any serious attempt to equate any two of them, much less all six.

On the specific concept of “information,” which Walker takes to be at the core of life, it is worth mentioning that there is little consensus on exactly what information is. Accounts divide between the broadly semantic, grounded in language, and more technical conceptions, generally based on the idea of reduction of uncertainty. Nowhere does Walker give a clear account of how she intends this term, and it is hard to see how an interpretation of either of these kinds will serve her ends.

I shall return, in conclusion, to the simplification of which the preceding quote is the most egregious example, but first I must mention a source of annoyance that has nothing much to do with my philosophical background. Walker adopts the curious convention of referring to all her sources by their first names after their first introduction. This is perhaps an understandable practice for people she knows or might have known, but I’m sure she didn’t know Charles (Darwin) or Albert (Einstein). This might just seem the complaint of an old-fashioned and stuffy academic, but it has its costs. On a few occasions, I couldn’t remember who Joe or Fred was. Most readers will have the benefit of an index, which should refer them to the first mention of a name. My prepublication copy didn’t have an index, so I had to trace back through the text to find the name. After one such excursion to find out the reference to “David,” I thought to run through the endnotes, which included (at least) seven different Davids. I assume the index will use last names, but if, bizarrely, it lists seven Davids under D, this will not easily resolve such questions. I am frankly amazed that a competent editor could allow this eccentric choice of style.

¤

Returning to more serious matters, the elision of concepts provokes a disturbing reflection on popular science generally. This is the, perhaps inevitable, value placed on simple, often simplistic narratives. Assembly theory may well be an illuminating way of thinking about many issues, but that it is the answer to the question of life, the universe, and everything is not remotely plausible. It makes for a good story, perhaps, and perhaps a story that will appeal to the nonspecialist reader.

This is not far from my deepest worry with Dawkins. The Genetic Book of the Dead, as with earlier representatives of the lineage, provides much that is edifying. Apart from interesting stories about animal behavior, there are many insights to be gleaned about the general shape of the evolutionary process. Fans of Dawkins the well-known atheist will find a number of telling points against some standard creationist arguments.

But the model of evolution that Dawkins continues to propound—selection between genes producing optimal adaptation—is no longer tenable. Dawkins does not, of course, think evolution is exactly simple. But he does think there is a simple principle, natural selection of genes, that is uniquely central to the process. Better understanding of development, epigenetics (very broadly construed), niche construction, genetic assimilation of phenotypic plasticity, symbiosis, and more provide major complications. Evolution looks increasingly like a complex narrative with different factors proving decisive at different moments, rather than the simple playing-out of a single natural law.

My discouraging suspicion is that the market for popular science, for whatever combination of public taste and commercial strategy, is strongly attracted to simple stories offering a quasi-theological insight into everything. Alternatives do exist. The book by Denis Noble, for instance, violently if ineffectively savaged by Dawkins, offers a sophisticated and properly complex picture of the evolution of life. I hope I am wrong, but I fear that such a book may struggle to have the popular success of Dawkins’s simple tale. There are, of course, some superb popular science books. But the temptation is ever-present to offer one deep insight that provides the illusion of a broader understanding than is truly attainable without a lot more attention to nuance and complexity.

Walker, and I suspect Dawkins, sees the world in a way that was widely accepted in much of the last century but is increasingly rejected by philosophers of science. This is a world in which, ultimately, the only truth is physics, and the only laws the laws of physics. Physical parts compose mechanisms, the behavior of which is determined by their physical parts, and which in turn may drive the behavior of more complex mechanisms. Dawkins does add a principle of natural selection to this, but it is a selection between physical structures, and these physical structures explain the behavior of the more complex organisms of which they are part. Organisms controlling the behavior of genes, the idea that Dawkins finds so offensive in Noble’s writing, is inconceivable in this world.

Walker is also firmly in this tradition. She even uses LEGO as an analogy for some of her arguments—an analogy often used by philosophers to ridicule the old, reductionist, and mechanistic worldview. The alternatives that she faces are either that life does not exist at all or that physics must be reconceived so that life is just a typical instantiation of the physical world. The latter is surely the preferable alternative, but a more wholehearted acceptance of the diversity of our universe would be even better.

I end with my own preferred way out of this philosophical dead end. The reductionist worldview is one firmly based on a traditional metaphysics, or ontology, that sees the world as composed of various kinds of things with their characteristic, sometimes essential or necessary, properties. But what biology, and I suspect even physics, is increasingly showing us is a world of process. An organism is not a thing with a fixed core of properties (e.g., the genome) but a process. It is a process constantly working to maintain itself, and following a trajectory from an initiating cell to, sometimes, a complex multicellular organism. This trajectory may pass through many phases, and the process of sustaining itself in an often-hostile environment must be more or less flexible and fluid. The complex and ever-changing relation to a multitude of symbiotic microbes fits easily into this picture. The recognition of the processual nature of our world might not be a key to all understanding, but it is an outstanding prophylactic against the simplistic interpretations too often promoted in popular science books.

¤

References

Dawkins, Richard. 1982. The Extended Phenotype. Oxford, UK: Oxford University Press.

Dawkins, Richard. 1996. The Selfish Gene. Oxford, UK: Oxford University Press.

Lewontin, R. C., and L. C. Dunn. 1960. “The Evolutionary Dynamics of a Polymorphism in the House Mouse,” Genetics 45 (June): 705–22. https://doi.org/10.1093/genetics/45.6.705. 

Méthot, Pierre-Olivier and Samuel Alizon. 2014. “What Is a Pathogen? Toward a Process View of Host-Parasite Interactions.” Virulence 5.8: 775–85. https://pubmed.ncbi.nlm.nih.gov/25483864. 

Noble, Denis. 2017. Dance to the Tune of Life: Biological Relativity. Cambridge, UK: Cambridge University Press.

Sober, Elliott and David Sloan Wilson. 1998. Unto Others: The Evolution and Psychology of Unselfish Behavior. Cambridge, MA: Harvard University Press.