The Loom

The Cost of Smarts

In tomorrow’s New York Times, I take a look at the evolution of intelligence. Or rather, I look at its flip side. Scientists and the rest of us are obsessed with intelligence–not just the intelligence of our own species, but any glimmer of intelligence in other animals. I’ve written plenty of stories myself on this research, from the social brilliance of hyenas to the foresight of birds. But if these faculties are so great, then why aren’t more animals smart? The answer, experiments suggest, is that learning and memory have nasty side-effects. They can even shorten your life (at least if you’re a fly).

This story has an odd back-story of its own. If you report on scientific research on evolution, sooner or later you’ll find yourself reading mind-blowing distortions of the science produced by creationists and people who make the same sorts of distortions and really really really don’t want to be called creationists. Sometimes they happen to choose some interesting research to distort, which, for me, is the silver lining in gloomy creationist clouds.

A couple years ago I discovered to my surprise that Ann Coulter devoted several pages in one of her books to misreading an article of mine about the appendix. Coulter couldn’t seem to understand that despite natural selection’s ability to produce adaptations, nature is filled with flaws (like my own defective appendix). One source of nature’s imperfection is the inescapable trade-off between the benefits some traits provide and the costs they incur. Coulter scoffed at experiments that suggested natural selection might not favor smart fruit flies. At about that point, I decided I had enough of Coulter and tracked down the original studies. I’ve been following this fascinating line of research ever since.


  1. #1 Pierce R. Butler
    May 6, 2008

    The downside to human intelligence also deserves more attention.

    If (as I’ve read but lack the time and energy tonight to find a linkable source) the artifacts found with early-human remains show no change during the periods when brain size was dramatically increasing, and if larger brains caused the greater maternal/infant mortality rate that modern obstetrics still contends against, then why wasn’t the big-brain experiment as much a failure in proto-hominids as it was with flies?

  2. #2 Edman
    May 6, 2008


    Perhaps the reason for the success of big brains in early apes and proto-hominids was a case of “right place at the right time”. In apes, the added intelligence could allow them to make tools, hunt in packs, etc. As for flies, I cannot imagine what intelligence alone would add, unless coupled with other features (such as dexterous hands) that would give the intelligence some use, and then make it a favorable trait.

    On a side note, Carl, I just want to say that my two favorite blogs here which address creationist claims are yours and PZ’s. I just read your entry about Ann Coulter, and while PZ’s “no-holds-barred” attitude makes many valid (and humorous) points, your attitude impresses me far more. To see you treat such a misrepresentation of your work without resorting to anything belittling or nasty speaks volumes about your character. Keep taking the high road, it makes a much better example.

    Oh, and I’m psyched to get my copy of Microcosm. Amazon says it’s in the mail.

  3. #3 JessC
    May 6, 2008

    Very interesting article. I wonder, however, if there are any visible biological feature differences between the short-lived-but-smarter fruit flies and the average ones, ie: more neurons, more connections?

    I’m tempted to extrapolate from the article that Ann Coulter would then outlive the rest of us, but she’s evil, not stupid.

  4. #4 Sven DiMilo
    May 6, 2008

    Nice article and fascinating subject. Neural tissue is indeed expensive to maintain in terms of energy: all those ion pumps and active reuptakers of neurotransmitters going all the time. It will be interesting to see if the proximate mechanisms underlying the learning ability / lifespan trade-off are more complicated than simple energy-allocation effects.

  5. #5 Dan Miller
    May 6, 2008

    I read the abstract, but they didn’t mention a mechanism–is this known?

  6. #6 agnostic
    May 6, 2008

    We focus on intelligence because that’s our own crowning achievement — but if we were flies, we’d be studying flight. It has the same rough features: incredibly expensive to maintain energetically, possessed to varying degrees by certain groups of animals and in somewhat different kinds, and most importantly is a way to deal with a highly variable environment.

    That’s true: in environments that are more variable, insects are more likely to have winged morphs. If the place you’re in right now sucks, you just fly somewhere else that’s better. Intelligence allows you to stay put but alter your environment so that it doesn’t suck so bad anymore (boiling poisonous plants, etc.).

    It’s like how some predators track prey by following electromagnetic gradients, others use chemical gradients, and others use motion-detectors and stereoscopic vision. They all get the job done, each probably being under strong phylogenetic constraint once it gets going. That could be why insects aren’t very smart: they’ve already gone so far down the path of “winged migration” as the solution to a variable environment.

  7. #7 Stacy
    May 6, 2008

    Very interesting! One thing occurs to me — independent of the energy requirements for bigger brains, a nonlearning society might just be more efficient making a living from the environment to which it’s adapted.

  8. #8 freds
    May 6, 2008

    Have you seen last year’s article by B. Crespi et al titled Adaptive evolution of genes underlying schizophrenia? They present evidence that schizophrenia is a side effect of selection for a large, intelligent brain. If Ann coulter had difficulty with intelligent fruit flies living shorter lives, the poor lady may really struggle while trying to think that human intelligence probably comes at the cost of 1% schizophrenia.

    From the Abstract:
    Schizophrenia poses an evolutionary-genetic paradox because it exhibits strongly negative fitness effects and high heritability, yet it persists at a prevalence of approximately 1% across all human cultures. Recent theory has proposed a resolution: that genetic liability to schizophrenia has evolved as a secondary consequence of selection for human cognitive traits. This hypothesis predicts that genes increasing the risk of this disorder have been subject to positive selection in the evolutionary history of humans and other primates.

    The link:

    Add this to the costs of painful, dangerous delivery and our large brains seem a mixed blessing.

  9. #9 daedalus2u
    May 6, 2008

    You should look at this article in Science. It shows that many genes in yeast are dispensable (~80%) while growing in the extremely protected laboratory environment with abundant food and trace nutrients.

    However something like ~97% of genes show measurable decreases in fitness following loss under more difficult environments.

    In fruit flies, the laboratory environment used was so simple, no predators, no wind, no temperature extremes that intelligence and learning doesn’t significantly aid survival or reproduction. There may be other interpretations.

  10. #10 Phred
    May 6, 2008

    Do you think there is a positive correlation between intelligence and “ready wit?” I’ve frequently noticed slow and somewhat pedantic persons make well thought out and insightful comments and judgments. And I have observed persons without any other seeming attributes except a “glib tongue” make successful politicians.

  11. #11 Colin Bartlett
    May 7, 2008

    As far as I’ve understood the arguments for our sudden ballooning of brain size, we got smarter primarily through social and sexual competition with each other, trying to mentally and politically outmaneuver our troopmates. As a means to this, we developed a mutation in typically conservative gene that helps to build the part of our brain used in mimicking fine motor skills, which in birds and bats is used to learn songs and sound patterns. A side effect of this was an increased complexity to our small motor skills, allowing us to do neat things like weave or tie complex knots (which many birds can do, but chimps cannot). While this particular aspect of our brain’s development would eventually lead to increasingly sophisticated tool creation, most of the social competition that inflated our brains would not. I actually wouldn’t expect our tool use to become markedly more complex until around the time we became pretty cunning linguists, which I personally expect happened with the evolution of H. sapiens itself, after most of the brainbuilding was finished.
    Of course, I’m just a sophomore undergrad at a public college, so I could be horribly misinterpreting all of this, and my apologies if I am.

  12. #12 Derek James
    May 7, 2008

    I’d actually like some clarification on these experiments, because I’m not sure I’m understanding it from the description in the article. As JessC said above, did they actually do any kind of comparative analysis of the nervous systems of each population of flies after the experiment, or did they just rely on behavior to infer that the flies learned more quickly.

    For example, in the first experiment Carl writes about, the quinine experiment, how do the researchers know that the adaptation is necessarily learning? Is it possible that the flies developed better “quinine detectors” (i.e. could it have been some kind of physiological adaptation to their sensory organs and not necessarily to their nervous system)?

  13. #13 Tony Jeremiah
    May 7, 2008

    The answer, experiments suggest, is that learning and memory have nasty side-effects. They can even shorten your life (at least if you’re a fly).

    Indeed there must be considerable variation in this effect across species. Research with humans suggests an opposite relationship, with greater learning and memory ability (as defined by IQ) corresponding with a longer lifespan. Specifically, Deary and Der (2005) showed that having a lower IQ significantly correlates with an earlier death, even controlling for education, social class, and smoking. However, this relationship becomes non-significant when controlling for reaction time. The authors argue that this is evidence that efficiency of information processing in the nervous system mediates the link between mental ability and lifespan.

    Given other research suggesting that intelligent human brains are also fuel efficient brains, perhaps there’s an opposite effect in flies, whereby an intelligent fly is a less fuel efficient one. This would seem to explain one of the larval competitive ability (LCA) results, whereby high and low learning flies did not differ in LCA when plenty of food was available. However, when less food was available, the high learning flies showed an LCA disadvantage.

    So being an intelligent fly might mean a “brighter” fly that burns out faster, whereas the opposite effect might occur for humans (if you exclude the stereotypical troubled artistic genius).


    Deary, I.J., & Der, G. (2005). Reaction time explains IQ’s association with death. Psychological Science, 16, 64-69.

  14. #14 Pierce R. Butler
    May 8, 2008

    Edman & Colin –

    Thanks for clear expositions of intriguing hypotheses.

    But I’m still of the opinion that there’s a larger mystery in this question (and no, I’m not setting up for a supernatural non-explanation).

    Apologies in advance for the late-night inelegance of my phrasing here, as well as for my raising a question probably sliced up and composted in the typical Paleontology 101 class.

    My response to both of your proposals: consider two groups of hominids, functionally identical except that one has a tendency toward larger-brained offspring. Members of the latter group may have an edge in terms of tying knots, more elaborate group interactions, etc: fairly subtle advantages, but the sort of thing that natural selection can work wonders with.

    However, the former group loses fewer females (and usually their infants as well) in giving birth – a rather unsubtle advantage, one that will lead to outcompeting the bigger-brained bunch by sheer population size in short order (assuming both groups start with at least an average primate level of social and physical skills).

    I.o.w: what (evolutionary) good is it for (say) a male to have a 1% advantage in attracting females, if his offspring have a 1.1% greater probability of killing their mothers and themselves at parturition?

    There aren’t enough fossils to put any kind of reliable numbers into this scenario, but the evident fact that women’s pelvises and birth canals are still so narrow that unaided birth can be highly dangerous suggests to me that the relevant prehistoric selection factors must have been intense, prolonged and unusual. (The pelvic-width issue in turn brings up the questions of our ancestral transition to bipedalism, a roughly contemporary process to that of brain growth and thus probably involving the same environmental conditions.)

  15. #15 APRudy
    May 8, 2008

    The questions about the experiments from daedalus2u and Derek are important but the wider arena is perhaps moreso.

    First of all, comparing the meaning of the “costs” (and presumably the “benefits”) of “intelligence” in fruit flies, early hominids and modern humans is to compare savanna grasses with pre-cultivation tomatoes and then genetically-modified anti-freeze tomato of the mid-80s. (I know this is imperfect but its better than the apples, oranges and tangelos I started out with.)

    In addition to the staggeringly problematic move of comparing the “costs” of “intelligence” to species as wildly different in their biogenetic make-up and sociality as insects, hominids and humans, and as folks with sociobiological tendencies oh-so-often do, this approach tends to abstract single or straightforward mutations and assume they relate to single rather than the uneven aggregation of traits and that these mutations and traits occur in straightforwardly bounded time outside of any relationship(s) with other “environmental” or “structural” factors whether, genetic, developmental, ecological or multi-trophic.

    Secondly, as sociobiologists are again so wont to do – and as Mr. Zimmer has uncritically reported in the NYTimes – it treats pregnancy and “unaided” childbirth as competitive and inherently and highly dangerous, while focusing on nutrient “competition” and skull/pelvic size. There is no question giving birth presents risks to mother and child. At the same time, there are overwhelming volumes of evidence-based medical research that show that the “aid” modern, masculinist medicine has provided women and babies has increased the dangers in childbirth rather than decreased them. Historically, well-fed, active and strong women – aided by other women – gave birth to children with what, by comparison to modern rates of complications, are “surprisingly” low complication, tearing and death rates (but note the stress on well-fed, active and strong women since our “modern” diets and lifestyles have, for at least two hundred years, produced large numbers of variously mal-nourished, physically pacified and weakened women.)

    Lastly, it is only if you take an asocial, reductionist and unsophisticated approach to the complexity of competition, predation, mutualism and commensualism (where symbioses can be facultative, obligate, oligophilic or polyphilic) that the issue of “costs” can be made so primary when it comes to the coevolution of human biology, sociality and transformation. Here, respectful or not, responding to the idealist and textual reductionism of creationists with the hyper-realist and economistic reductionism of modern biologism does science, its richness, its material complexity and its beauty few favors… or at least that’s how I see it.

    Sorry for the screed.

New comments have been disabled.