Darwinian Reductionism, Or, How to Stop Worrying and Love Molecular Biology must be one of the toughest books I’ve read. Luckily, it’s fantastic.
Alex Rosenberg is a professor of both Philosophy and Biology at Duke University. He has written about the philosophy of science before, like a book about the non-validity of economics as a science. Rosenberg is a true intellectual powerhouse, and to watch his mind work over the course of this book’s 238 pages (+ about 30 pages of references and index) is one of the pleasures of reading this book.
Kim Sterenly sketches what it’s about on the back cover:
“Over the last twenty years and more, philosophers and theoretical biologists have built an antireductionist consensus about biology. We have thought that biology is autonomous without being spooky. While biological systems are built from chemical ones, biological facts are not just physical facts, and biological explanations cannot be replaced by physical and chemical ones. The most consistent, articulate, informed, and lucid skeptic about this view has been Alex Rosenberg, and Darwinian Reductionism is the mature synthesis of his alternative vision. He argues that we can show the paradigm facts of biology – evolution and development – are built from the chemical and physical, and reduce to them. Moreover, he argues, unpleasantly plausibly, that defenders of the consensus must slip one way or the other: into spookiness about the biological, or into a reduction program for the biological.”
But for many people, including scientists, there are problems with materialistic reductionism, as Elliot Sober explains on the back cover, before pointing out how Rosenberg tackles those problems.
“For most philosophers, reductionism is wrong because it denies the fact of multiple realizability. For most biologists, reductionism is wrong because it involves a commitment to genetic determinism. In this stimulating new book, Rosenberg reconfigures the problem. His Darwinian reductionism denies genetic determinism and it has no problem with multiple realizability. It captures what scientific materialism should have been after all along.”
I will not get into the nuts and bolts of every argument. Aside from a general appraisal of the book, I’ll elaborate a bit on two small – yet fundamental – elements of critique, and end with a list of nuggets of wisdom I found while reading – a list that is probably of interest to those readers not interested in the general content of this book, yet who do have a healthy interest in science.
Darwinian Reductionism is fairly self-contained, in that it spells out every argument in clear terms. One doesn’t necessarily need a lot of working knowledge about molecular biology or philosophy, but you better have a basic command of some vocabulary: basic biological or philosophical terms like allele, nucleotide, mitochondrion, epigenetics, ontology and etiology are not explained. There’s dictionaries for that, should you need them. Highly specific terms to some of the arguments (like how some molecules “supervene” in a philosophical sense, or the difference between a “proximate” and an “ultimate” explanation, a distinction first made by Ernst Mayr) are explained.
Nonetheless, this book has a limited audience. It’s not meant for laypersons – a popular science book this is not – but rather for biologists, philosophers with an interest in science, scientists with an interest in philosophy, and, more generally, any intellectual with an interest in the fairly specific subject matter.
As I’ve always been interested in evolution and problems concerning scientific materialism – more specifically anything that deals with the causality of our own actions – I enjoyed reading this immensely, even though big parts of chapter 5 and 6 (about 30 pages in total) were a bit too technical for me to really fully understand.
Rosenberg writes precise, clear prose. It’s not bloated nor pretentious, and has an honest approach. Big parts of the book are devoted to carefully examining the arguments against macromolecular reductionism, and it is a cerebral joy to see every argument followed to its conclusions, based on the available evidence. Rosenberg uses both philosophy and science to frame the discussions. Needless to say, there’s lots to learn on both fronts, as Rosenberg draws from a vast array of sources, smoothly incorporating insights from, let’s say, the philosophy of history, or new technological developments in gene sequencing.
The fact that the book was published in 2006 obviously makes it not fully up to date anymore – molecular & genetic biology evolving rapidly – but that doesn’t hinder Rosenberg’s arguments, as the overall debate runs much deeper than the latest cutting edge news in Nature. Coincidentally, I’ve read a much more recent book about molecular biology and genetics a few months ago – The Vital Question: Energy, Evolution and The Origins of Complex Life by Nick Lane (2015) – and as far as I can see current evolutions in the field only validate Rosenberg. The only thing I spotted that was clearly outdated was the claim Neanderthals didn’t interbreed with Homo Sapiens, but again, that fact doesn’t run against his main arguments.
While I don’t claim to be a specialist nor a philosopher, there are two issues I want to bring up.
1. While I agree words matter, a part of chapter 8 (the final chapter) actually revolves around semantics: there is a difference between ‘causation’ and ‘determinism’, but it is crystal clear that PKU (an inherited metabolic disease) is caused by some combination of genes. The fact that one can alter the environment to deal with this disease (change one’s diet) does make these genes less (or not fully) determining in one sense. True, genes don’t necessarily cause the disease as one can change one’s diet, but in that case, the genes do cause the need for dietary changes. In that respect, there still is determination.
Similarly, the fact that not merely one gene or group of genes causes this or that effect, but that often different genetic pathways also lead to a similar outcome, doesn’t make the overall statement that genes determine less true. It only refutes a very narrow approach to genetic determinism. A part of Rosenberg’s refutation of genetic determinism boils down to ‘it’s too complex to describe’; but that’s obviously not a valid argument.
All things considered, it seems that Rosenberg is fighting a specific, very narrow definition of genetic determinism. It’s obvious that genes are not the full story, but they – and all the molecular structures related to them – irrefutably do their part in materialistically determining biological outcome.
2. I missed Rosenberg’s stand on free will. It is the elephant in the room that isn’t addressed at all in this book. In chapter 8, on human behavior, Rosenberg only talks about genetic determinism, while in much of the rest of the book, he often talks about the more general macromolecular reductionism. Why not go the full mile, and write a chapter about the consequences of general material determinism? Or is the jury still out? He sure didn’t go there out of fear for controversy, as the rest of the book is also controversial: a large part of biologists apparently are physicalists yet anti-reductionist.
Opponents of material reductionism seem to be frightened that it will lead to some kind of genetic determinism, as Rosenberg writes. But I think the real fear looms even larger, and that fear explains the absence of a discussion of free will. So it’s not only about some conceptually narrow form of genetic determinism, it is about a fear for full blown material determinism: scientific materialism, and what that might mean for consciousness, theory of mind and, ultimately, free will and parts of most moral systems.
Rosenberg’s opinion seems to seep through at times, when he says things like the body’s responses are “the consequence of the operation of a structured program of subroutines”. Yet Rosenberg doesn’t seem immune for that fear himself. On page 228 Rosenberg warns against “the slippery slope to full-blown and morally unacceptable genetic determinism of socially significant traits.” It’s morally unacceptable because of possible “racist, sexist or other nefarious conclusions”. I can sympathize with that, but I still have problems with “morally unacceptable”. It seems strangely out of place as an argument in what is in essence an ontological question.
If you’re interested in the matter, there’s an excellent essay by Anthony Cashmore I’ve linked to before on this blog a couple of times, The Lucretian swerve, a text that imho settles the matter. Not for the fearful!
Now for that list of nuggets of wisdom:
> The crucial difference between biology on the one hand, and physics and chemistry on the other, is that biology is a fundamentally historical science. When we talk about human biology for instance, we talk about something very specific, and even very limited when considered on the geological timescale. As such, there are no fundamental biological laws – except for the principle of evolution. DNA is an historical trend, not a fundamental law. (I should note Rosenberg fails to mention that also the laws of physics and chemistry haven’t existed for ever, but obviously he is right when we limit ourselves to the current universe of stars and planets.)
> Likewise, biology is terrestrial, chemistry/physics have a larger field of application. (But it would be interesting to take the conclusions of that Nick Lane book into consideration here, as he makes the claim that the way life might have originated on earth is maybe the only way possible from a molecular/energetic point of view, so that evolution and biology would have very strong parallels on other planets or in other non-terrestrial ecologies.)
> Evolution selects for effects, not physical structure. E.g. there are countless biological ways to make wings. Similarly, the eye has evolved independently multiple times in very different ways, but the genes that start the embryonic process of building eyes go all the way back to the beginning of the common ancestor of both the fruitfly and the mouse. That’s why variation is the norm in biology, and what consists as ‘normalcy’ is just something statistical, not something essential or typological. So in biology, there are no real generalities, only natural history: the product of the laws of natural selection on macromolecular initial conditions.
> In the same way, chemistry/physics is about a structural taxonomy, biology about a functional taxonomy.
> Both the political right and the political left have an inbuilt trepidation for material reductionism. On the right reductionism might shed light on the question whether inequality is morally permissible or not, while on the left they oppose the possibility of factors like race, gender, etc. as determinants.
> A Wittgenstein snippet I didn’t know or had forgotten: “anything can be similar to anything else, and criteria of similarity reflect our interests.”
> A good refutation of Putnam’s “square peg-round hole” argument, that states that a geometrical explanation of why square pegs don’t fit round holes is always a better explanation than a (quantum)mechanical one. Rosenberg argues that once you take the context of the peg & hole and their substance into account, mathematical geometry fails, and needs fundamental physics. Consider a square peg made of soft jelly and a round hole made of stone.
> There’s lots and lots of stuff about genes we don’t understand at the moment. It’s way too complex, and …there even is no exact definition of what accounts as a ‘gene’!
> Natural selection is also at work in chemistry, but it’s so obvious, that it’s hardly ever noticed: stable molecules in a certain environment (molecules with an apt fitness) are being selected for.
> I’m not sure if this holds up to current scientific insights, but Rosenberg notes that until about 12,000 years ago, humans probably lived more in stable pairs than in larger groups.
> Cooperation among humans evolved and spread so rapidly that it is probably not explainable as the sudden result of a genetic mutation.
I’ll end with 2 anecdotes that say something about the baffling complexity of evolution.
> Certain genes of male parentage and certain genes of female parentage have different effects on maternal fitness. There’s a parental gene that “helps produce a strong placenta which advantages the fetus carrying the paternal genes at the expense of the mother, since better nutrient transfer weakens her ability to survive into future breeding seasons.” The maternal gene “limits placenta development, with fitness-enhancing effects for the mother and her genes. This will be especially true in breeding systems such as that of the mouse, where a single litter may have mixed paternity, and there is little likelihood that the mother’s future offspring will share paternity. (…) a classic case of genomic conflict, of an arms race at the level of the polynucleotides”. The case is a bit more complex, but I’ll leave it at this.
> A male mouse tends to kill any mouse pups it encounters, aside from its own offspring. Yet technically, that’s not really the case, as the mouse doesn’t even recognize its own offspring. It’s just that he evolved a “package of genes that automatically switches off the mouse’s pup-killer disposition from day 18 to day 22 after its last ejaculation. This period happens to be the gestation time for female mice. So, the pups the male encounters during this period have a high probability of being its own pups”.
Last year, Rosenberg published a book about the implications of neuroscience for the nature of history. Here’s that book’s page on MIT’s website, and here’s an interview with Rosenberg on The Verge. Worth a peek.
I’ve ordered a copy even before I finished Darwinian Reductionism: I think I’m a fanboy. (update April 2020: here’s my review of How History Gets Things Wrong.)
Click here for an index of my non-fiction or art book reviews only, and here for an index of my longer fiction reviews of a more scholarly & philosophical nature. Here’s a link to my list of favorite non-fiction books.