It may be a while before I start reviewing two publications a week, but maybe reviewing books should count for more? I recommend both:

“Force of Nature: Understanding Evolution’s Deepest Logic — and Putting it to Use” by Owen Jones.

“Evolution Gone Wrong: the Curious Reasons Why our Bodies Work (or don’t)” by Alex Bezzerides,

I just finished reading Force of Nature, whereas I listened to Evolution Gone Wrong as an audiobook last year, so I’ll say more about the former.

Force of Nature starts by explaining why natural selection was/is more powerful than many people realize (it’s had longer to work, can be faster, etc.). That leads to a major theme, that ignoring natural selection is a really bad idea. For example, the fact that infection triggers fever in so many species — “cold-blooded” ones move into warmer spots when infected — suggests that it may be an evolved defense, rather than something we should always try to reverse. This was something I’d read about in Why We Get Sick which I’ll review some time. By coincidence, I’d just been reading The Confusion, in which a character has syphilis cured by the fever of malaria. Not a recommended cure today!

Next comes a chapter on implications of natural selection for our food supply. The agriculture part draws heavily on my book and probably on some email exchanges I’d had with Jones. Then there’s a section on how we’re imposing harmful selection on ocean fish populations, by using nets that selectively remove fish that mature later and grow larger.

After some interesting material on evolutionary computation and biomimicry, he moves on to his main interest (and more material that was new to me): evolution of behavior. I knew about archer fish, but learned that they accelerate their water jet so that it arrives all at once, increasing the chance of knocking their insect prey off a branch.

A major subtheme is humans often tend to behave in ways that, shaped by past natural selection, are not optimal in current environments. Many people are familiar with the hypothesis that “eat as much fat as you can” was often a fitness-enhancing strategy in the past, but isn’t in many countries today. But Jones presents two additional examples of “time-shifted rationality” that were new to me.

First, you may be aware that most people are terrible at conditional-risk calculations, such as “if 0.1% of a population has disease X and the text for disease has a 10% false-positive rate, what percent of people with a positive test actually have the disease?” Apparently, many doctors assume that if the false-positive rate is “only” 10%, most people who test positive have the disease.

But of 1000 people tested, the test would correctly identify the one person with the disease (assuming no false negatives) plus 100 false-positives (10% times 1/1000). So only 1% of those who test positive actually have the disease. (Similarly, people who buy ammonium nitrate are more likely to be making bombs than people who don’t, but most just want if for fertilizer. And men are more likely to commit murder than women, but most men don’t commit murder.)

Where does evolution come in? Apparently some evolutionary biologists came up with the hypothesis that our ancestors only dealt with integers (3 of 10 people who are the berry died), not decimals (0.3) or percents (33%). They found that presenting the same problem using integers increased the fraction getting it right by several-fold.

A final interesting example was the “endowment effect”, like when someone who would only pay $1 for something won’t sell it for $2 once they have it. This apparently causes all sorts of “transaction costs” and problems with government policy, like opposition to single-payer plans, even if they would replace employment-based health insurance with better coverage for less money. Jones didn’t use that example, but did describe experiments with chimps in which he was involved, which showed an unwillingness to trade “a bird in the hand” for a better one “in the bush.”

I highly recommend this book. I didn’t spot any significant errors, just thought of a lot of things that could have been included, but weren’t. Some (many?) of those things are in the extensive bibliography, which could keep one busy for years. And limiting the length (and hence, number of examples) to something that could be read over a weekend was probably a good idea.

Aside from specific examples, like Jay Rosenheim’s great essay making an evolutionary argument for the feasibility of curing most cancers (cited), what’s missing?

The opening chapters emphasize the power of natural selection to solve problems. Jones mentions once or twice that we shouldn’t expect perfection (partly due to tradeoffs) but the main discussion of natural-selection’s limits is of cases where natural selection got it right, but it’s not right any more, which he calls “time-shifted rationality” in the context of newly-maladaptive behaviors. But that leaves out a whole class of cases where natural selection will never get it right (by human criteria), even if conditions stay constant for millions of years.

These evolutionary “mistakes” include Prisoner’s Dilemma or Tragedy of the Commons situations, where genuinely optimizing individual fitness leads to bad consequences for the population as a whole. Like Irish elk going extinct because their giant antlers got stuck in trees while, at the same time, males with bigger antlers got all the mates. (There’s some doubt about this story, but it may be more memorable than “corn tassels made much more pollen than the crop needed, collectively, while shading lower leaves enough to reduce photosynthesis and collective seed production — and it took 60 years for breeding-for-yield to solve this problem.”) The “discounted-legacy” problems students are working on in my lab — crop genotypes that improve conditions for future crops, e.g., by favoring more-beneficial symbionts, probably pay a small fitness cost, so traits that would optimize crop production won’t evolve without our help — are in some sense a variation of this Tragedy of the Commons problem. For human examples see Robert Frank’s The Darwin Economy.

Anyway, great book! Moving on…

I should go back and reread Evolution Gone Wrong, but here’s some of what I remember. Yes, we’ve inherited bodies that were shaped by evolution in very different environments, but that’s not the only problem. There are tradeoffs that even a superintelligent designer — let’s call him “Claude” — would not be able to overcome. Apparently (within some design space), you can either have the ability to make speech-like sounds, or be immune from choking, but not both. Chimps and newborn humans can’t choke, but they can’t talk either. Then (in humans) the voicebox moves to the speech-synthesis position, which is an impressive adaptation, but accepts the choking-susceptibility tradeoff.

Then there’s the you-can’t-get-there-from-here problem. A different genotype might be better in a new environment, but evolution can only get there in a series of steps, with each step being viable. As I recall, the book holds out the possibility that our eyes might eventually be optimized for air, but so far they retain some traits that helped our fishy ancestors to see under water. There’s lots of material on pregnancy, some of it also involving tradeoffs that our descendants will probably have to live with indefinitely, barring some Brave New World or Woman on the Edge of Time technological breakthrough.

Anyway, I recommend this book by Alex Bezzerides (possibly a pseudonym for Oolon Colluphid).