Book review: Improbable Destinies by Jonathan Losos

Jonathan Losos is one of the world’s most eminent evolutionary biologists. Last year he published his first popular science book, Improbable Destinies: Fate, Chance, and the Future of Evolution. Here’s my review. For other reviews, see The Molecular Ecologist, Science, and Goodreads users.

Back in 1989, the late Stephen J. Gould published Wonderful Life, using the bizarre Cambrian animals fossilized in the Burgess Shale to argue that the history of life on Earth is highly contingent. If you “rewound the tape of life” and played it again, evolutionary history would come out completely differently. Gould’s conclusion was controversial. Perhaps his best-known opponent was fellow paleontologist Simon Conway Morris, who argued that in fact the Cambrian Explosion was one of many examples illustrating evolutionary convergence: that evolution by natural selection predictably and repeatedly evolves similar species in response to similar selection pressures.

Improbable Destinies is Jonathan Losos’ updated look at Gould vs. Conway Morris. That different experts could interpret the Burgess Shale fossils in opposite ways suggests that fossil evidence can’t really decide the question of chance vs. determinism in evolution. If you want to know how often a given process will come out one way vs. another, the most obvious thing to do is to try it many times. That is, run a replicated experiment. Losos’ book reviews lab and field experiments testing the repeatability of evolution.

He starts with his own work on anoles in the Carribbean, which provide many textbook examples of convergent evolution. I was intrigued to learn that, before starting this book, Losos had been a Gouldian, considering Carribbean anoles to be an exception to the general rule that evolution isn’t repeatable. I wonder how many leading researchers consider their own study systems to be “exceptions to the rule”? Not many, I bet. Anyway, Losos then moves on to other evolution experiments, many of them quite famous in the field: Trinidadian guppies, threespine stickleback in postglacial lakes, the Long-Term Evolution Experiment, the Pseudomonas model adaptive radiation, Park Grass, and others. He also discusses various famous “natural” experiments, such as the repeated evolution of antibiotic resistance in bacteria and pesticide resistance in insects.

If that list of examples mostly sounds familiar, well, that may mean that you, like me, aren’t really the target audience for this book. I do sometimes enjoy reading about science and scientists already familiar to me. But I have to learn something new–entertaining new anecdotes, a fresh perspective on a familiar topic, etc. For a reader like me, who’s read a bunch of papers from folks like Rich Lenski, Paul Rainey, Dolph Schluter, David Reznick, and Losos himself, too much of Improbable Destinies was familiar. I often found myself able to see too far ahead, like someone reading a thriller who spots the plot twist coming well before it arrives. For instance, a large chunk of one chapter is framed around the difference between the repeatability of evolution starting from exactly the same initial conditions, vs. the repeatability of evolution starting from slightly different initial conditions. I’m sure that framing is fine for a general reader, but I immediately thought, “Wait a sec, that’s a distinction without a difference. Even if you start an experiment with identical replicates, slight among-replicate differences are going to arise soon enough. At which point different replicates could be said to have slightly different initial conditions.” I found myself skimming to get to the point where Losos says as much himself.

The last chapter is a departure from the rest of the book, as Losos muses on whether evolution of humans, or something like them, was inevitable, and whether life on other planets would resemble life on Earth. This was fun and astute, but also ungrounded in anything resembling data. But quite possibly, the general readers at whom the book is aimed might not notice, or might not mind, the contrast with the experiments comprising most of the book.

The style is conversational, bordering on folksy. As the Molecular Ecologist review notes, Jonathan Losos loves dad jokes and corny alliteration. He also gushes about the scientists he’s writing about. For instance, Losos writes in a joke footnote that Rich Lenski isn’t omnipotent “as far as I’m aware”.

The main conclusion I took away from the book is that somebody needs to figure out how to get beyond case studies of chance vs. determinism in evolution. Don’t get me wrong, I love well-worked out case studies that test some general hypothesis. The experiments Losos summarizes include many of my all-time favorites. And I’m a quality over quantity guy; I would much rather have a few really good case studies that can only be summarized and compared qualitatively, than a whole bunch of mediocre ones that can be summarized and compared quantitatively (e.g., via meta-analysis). But of course, many good case studies would be better than fewer. I’m thinking back to Rees Kassen’s Experimental Evolution and the Nature of Biodiversity, the core of which are meta-analyses of microbial evolution experiments addressing various fundamental evolutionary questions. It would be super-cool if over the next couple of decades evolutionary biology built up a critical mass of experiments on chance vs. determinism, so that the next book about that subject could provide a definitive, quantitative answer.

In summary, it’s a very good popular science book. It’s framed around a big idea and summarizes what a lot of really outstanding science reveals about that idea. As long as you are part of the target audience, I’m sure you’ll enjoy it. But it might be a bit slow for readers who already know a lot about the topic.

3 thoughts on “Book review: Improbable Destinies by Jonathan Losos

  1. Hmmmm, sounds interesting, I’ll have to put it on my Christmas list. However my immediate reaction from your review is that Losos and Gould are not talking about the same thing. Gould’s metaphor of “rewinding the tape of life” was focused on macroevolution of body plans and other major changes in form, over very long timescales (i.e. rewidning back to the Cambrian). Whereas the examples you cite above are of microevolution over rather short timescales.

    Also, my reading of Gould is that convergent evolution does not falsify his metaphor: placental wolves and the Tasmanian “wolf” may look superficially similar, but they are fundamentally different types of animal with rather different reproductive biology, etc.

    • “Gould’s metaphor of “rewinding the tape of life” was focused on macroevolution of body plans and other major changes in form, over very long timescales (i.e. rewidning back to the Cambrian). ”

      Yes, Losos notes that in an interview about the book.

      “placental wolves and the Tasmanian “wolf” may look superficially similar, but they are fundamentally different types of animal with rather different reproductive biology, etc.”

      The book does get into that–how you can have convergence in some traits but not others, or phenotypic convergence but not convergence in the underlying genotypes, etc.

  2. Ard Louis has a few recent papers on models of convergence in evolution, and how these fit in with DNA transcription and organization etc. This one in particular is nice, and uses some toy models to explain some aspects of convergence. I know this is quite a different approach from case studies and natural history/palaeontology, but perhaps such things form a nice complement to this experimental work.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.