Are humans innately bad social scientists?

I know, this sounds a bit extreme. How can the ability to do (bad) social science be influenced by our genes? Well, quite easily if you carefully read Robert Trivers’ last book (see reviews in NYT Nature, Science). Indeed, his book is about our innate tendency for self-deception. Here is the blurb:

Whether it’s in a cockpit at takeoff or the planning of an offensive war, a romantic relationship or a dispute at the office, there are many opportunities to lie and self-deceive—but deceit and self-deception carry the costs of being alienated from reality and can lead

In his bold new work, prominent biological theorist Robert Trivers unflinchingly argues that self-deception evolved in the service of deceit—the better to fool others. We do it for biological reasons—in order to help us survive and procreate. From viruses mimicking host behavior to humans misremembering (sometimes intentionally) the details of a quarrel, science has proven that the deceptive one can always outwit the masses.todisaster. So why does deception play such a prominent role in our everyday lives? In short, why do we deceive?

Among all the fascinating consequences of the evolution of self-deception – false memory, parents-offspring conflict, space disasters – one is of particular interest for us here at the ICCI. It is our innate propensity to do bad social science.



The social sciences, argues Trivers, are intrinsically difficult to practice because they require for people to go beyond all the little social lies we tell ourselves to achieve social success. This observation leads Trivers to propose a sort of ‘law of the failure of social sciences”:



“the greater the social content of a discipline, the more slowly it will develop because it faces, in part, greater forces of deceit and self-deception that impede progress.”

He goes on:

“Thus psychology, sociology, anthropology, and economics have direct implications for our view of ourselves and of others, so one might expect their very structure to be easily deformed by self-deception. The same can be said for branches of biology, especially social theory and (separately) human genetics.

Trivers starts with biology:

For roughly a century, biologists had the social world analyzed almost upside down. They argued that social selection favored what was good for the group or the species, when in fact it favors what is good for the individual (measured in survival and reproduction), as Darwin well knew. More precisely, natural selections works on the genes within an individual to promote their own survival and reproduction, which is usually equivalent to what is beneficial for the individual propagating the genes. Yet, almost from the moment Darwin’s theory was published, scientists in the discipline reverted to the older view of benefit as serving a higher function (species, ecosystem and so on) and only now they cite Darwin as support for their belief. In turn, the false theory was just the kind of social theory you would expect people to adopt in group living species whose members are concerned with increasing one another’s group orientation. (…)

For example, take the classic case of male infanticide, first studied in depth in the langur monkeys of India, and now known for more than one hundred species. Male murder of dependent offspring (fathered by previous males) was rationalized as of a population-control mechanism that kept the species from eating itself out of house and home. Male murder thus served the interest of all. Of course, it did no such thing. Since a nursing infant inhibits its mother’s ovulation, murder of the infant brought the bereaved mother into reproductive readiness quicker, which aided the male’s reproduction but at a cost to the dead infant and its mother. In some population, as many as 10 percent of all young are murdered by adult males—each murder gaining on average only two months of maternal time for the new male.

Note that the error is virtually irrelevant for non social traits. The human locking kneecap allows us to stay erect without wasting energy in tensed legs. It evolved because it benefited the individual with the new kneecap, but if you said it evolved to benefit the species, you would not misinterpret the kneecap. Not so for social traits. Here, as we have seen, we can exactly invert the meaning of a trait by failing to see how it is favored among individuals, even though it may be more costly to others. Instead, we imagine that everyone benefits. This often amounts to reaffirming Pangloss’s theorem—that everything is for the best in the best of all possible worlds.*

Trivers then consider the case of economics (“Is economics a science? The short answer is no.”**).

And eventually, anthropology:

Social anthropology made a tragic left turn in the mid-1970s from which it has yet to recover (at least in the United States). (…) In the early 1970s, a strong social theory emerged from biology and a variety of subject were addressed seriously from the first time: kinship theory, including parent/offspring relations, relative parent investment, and the evolution of sex differences, the sex ratio, reciprocal altruism and a sense of fairness, and so on. Social anthropologists had a choice: accept the new work, master it, and rewrite their own discipline along the new lines, or reject the new work and protect their own expertise (such as it was). As one has been noted “Faced with the choice between changing one’s mind and proving that there is no need to do so, almost everyone gets busy on the proof.”

Certainly social anthropologists rose to the challenge, even renaming their field ‘cultural anthropology’ to more explicitly rule out the relevance of biology in advance. Now we were no longer social organisms but cultural ones. This justification, in turn, was moral. Out of biological thinking flowed biological determinism (the notion that genetics influences daily life), whose downstream effects included fascism, racism, sexism, heterosexism, and other odious “isms”. (…) Thus an entire new area of social theory would be ruled out based on the alleged pernicious influence of its assumptions, which were, in fact, widely accepted as true (genes exist, they affect social traits, natural selection affects their relative frequencies, and this produce meaningful patterns).

In a way, Trivers’ law is not new. One could find it for instance in Bourdieu’s work (or in the famous Bourdieu-centered movie “La sociologie est un sport de combat”). For Bourdieu, sociology was intrinsically difficult because it has to fight against dominant social forces and symbolic dominant narratives. But I think there is something new and fresh in linking the difficulty of doing social science to a more general phenomenon that also arises in daily life, and even among animal and plants.

The passage on biology is also very interesting. First, it suggests that the problem is not about social sciences per se (the big business and so on), but about any kind of theory that goes against our entrenched interest in seeing us as disinterested good people. At a time when many people try to understand why Americans are set up against evolution, it may be interesting to keep in mind that American’s reluctance to embrace evolution is not only due to the intuitiveness of teleologism, or the influence of religion, but also to the fact that nobody wants to know that one is the product of an unfair and selfish process of mutation and selection. The fact is, even in secular France, where no one denies the existence of evolution and where the history of life is taught in every high school, a biologist like Gould who focuses on the history of life, its contingencies and little miracles is much more popular (among lay people, but also among scientists and intellectuals) than a biologist like Dawkins who portrays us a “as gene machines, created to pass on our genes”.

Finally, the bitter note on anthropology raised an interesting question, especially at a time when behavioral economics, cognitive neurosciences, experimental psychology and evolutionary biology are converging toward a new anthropological paradigm outside of the academic discipline of anthropology. What went wrong thirty years ago? Why didn't anthropology become the central point of this convergence? When one consider Trivers’ pioneering work on reciprocity, self-deception or parent-offspring conflict or when one re-read the account of the first meetings on this topic at the AAA in Evolutionary Biology and Human Social Behaviour, one can see that everything was already in place for the emergence of this new paradigm. Was it a moral problem as Trivers suggests? Or was the problem more epistemologic? Maybe the incorporation of evolution in social science was not possible before the proximate mechanisms were better understood? Or maybe we just needed a critical mass of “naturalist anthropologists”? But then, why did anthropology suffer so much while psychology and economics flourished?

What do you think, dear readers? Was academic anthropology killed by self-deception? And also: are we really innately bad social scientists?


*I can’t resist quoting his opinion of experimental games:

"One recent effort by economists to link up with allied disciplines is called behavioral economics, a link with psychology that is most welcome. But as usual, economists resolutely refuse to make the final link to evolutionary theory, even when going through the motions. That is, even those economists who propose evolutionary explanations of economic behavior often do so with unusual, counterlogical assumptions. For example, a common recent mistake (published in all the best journals) is to assume that our behavior evolved specifically to fit artificial economic games."

** “What are we trying to maximize? Here, economists play a shell game. People are expected to attempt to maximize their ‘utility’. And what is utility? Well, anything people wish to maximize. In some situations, you will try to maximize money acquired, in others food, and in yet others sex over food and money. So we need “preference functions” to tell us when one kind of utility takes precedence over another. These must be empirically determined, since economics by itself can provide no theory for how the organism is expected to rank these variables. But determining all of the preference functions by measurement in all the relevant situations is hopeless from the outset, even for a single organism, much less a group.


  • comment-avatar
    Olivier Morin 26 January 2012 (13:15)

    Just to sound a more positive note… Judging from what you quoted, Bob Trivers seems to voice a common complaint among scientifically-minded people. To put it bluntly, all the real progress would come from the natural sciences. Social science would be a backwater of academia, because it is crippled by political prejudice. This seems quite unfair to me. The criticism forgets at least two things.

    First (I know we agree on this point, Nicolas) there was progress in the social sciences. Two hundred years ago, our understanding of human history was barely more than official chronicles with a propagandist spin (or a philosophical one if we were lucky). Most people thought that human races were distinct species, or children of Adam, innately incapable of reaching the same levels of progress. The idea of a secular morality was for the privileged only. There was almost no economic history, no ethnography, no psychology. The inferiority of women was a given. Statistics were just being born, thanks to demographers, other social scientists, and mathematicians (even the mighty physicists took some inspiration from the results).

    Why were those progresses less visible than those of the hard sciences? In my view, that was because they were rapidly appropriated, criticized and transformed by people outside the social sciences. Statistics, feminism, network theory, experimental approaches to poverty — and evolutionary psychology too! — are enthusiastically adopted outside the academia. In contrast, the theoretical achievements of the harder sciences are too arcane to get that kind of reception.

    This porosity between social science and civil society is, I think, perfectly predictable. Because Trivers is wrong. Humans are excellent social scientists. (The best social scientists I know, it recently occurred to me, are all humans. Funny.) Our cognition evolved to fit the harsh requisites of a Machiavellian society; we deal with complex relations and institutions every day of our lives. We would not survive if we were not experts. Because they are surrounded with such a demanding audience, social scientists cannot claim the same kind of expert status as natural scientists. As a result, they cannot reach the top of the academic pecking order. Pity us.

    But there is an upside. We have access to a reactive, vibrant and knowledgeable civil society. (Hard scientists often complain that they don’t.) This is a point that anthropologists have perfectly understood, and many have made it the basis of their work. This has yielded some outstanding achievements (yesterday, Franz Boas’ struggle against racism; today David Graeber’s work with the anti-globalization movement). Yes, anthropology should not be about advocacy, activism and nothing else. Yes, there should be room for theory, evolutionary theory in particular. Yes, there should be more consilience. Yet it seems to me that we should start by acknowledging the value of anthropology and other social sciences.

  • comment-avatar
    Gordon Ingram 26 January 2012 (14:52)

    Olivier makes a good point. Social scientists don’t get much respect from laypeople because we share similar domains of interest to them. We are perceived as simply putting into fancy words the kind of thinking that the folk do intuitively (perhaps we are even sometimes thought to have the conscious aim of mystifying them!). In contrast, the domains of interest of physical scientists are intuitively opaque to most folk, so the physical scientists are perceived as brainy.

    It’s a bit like the comment, which always infuriates me: “Oh so you’re not a real doctor – you only have a PhD.” I usually smile back at them through gritted teeth, but some day they’ll catch me in a bad mood and I’ll snap back, “Yes, my doctoral degree involved crafting and reporting a substantial, original piece of research, rather than the rote learning and regurgitation of medical facts!”

  • comment-avatar
    Radu Umbres 28 January 2012 (00:32)

    From an anthropological corner, I agree with Olivier that humans make good social scientists. In fact, humans are excellent social scientists of their own society – competent users of a folk theory of the social which would include things like a model of a person, a theory of social relations, kinship, etc. (I believe Maurice Bloch suggested this in an interview).

    (Moreover, these theories would hardly go wrong since they could be considered as self-fulfilling prophecies – if I think that the nature of social relationships in my society is X, I will behave, as a competent social being, in line with this theory, bringing about the reality predicted by the theory. )

    One of the fundamental aims in ethnography is to reproduce (in indirect speech) a translation/interpretation of these folk theories. Given the complexity of beliefs, the variations of theories within a population, the methodological constraints of (usually) a single fieldworker and various sundry problems, this task is very difficult. Taking this into account, anthropology has largely succeeded in showing us the so-many ways of being human.

    This success is even more laudable since the wider audience was not only taxing (as competent users of their own folk social theories) but sometimes even opposed to the entire project. Olivier mentioned the work of Boas in fighting against racism (and there were many more) – just for historical comparison, I don’t think Rutherford or Galton were working against a tide of public opinion(I dare say quite the opposite).

    Having said this, I am also inclined to think that folk social theories (and their corresponding ethnographies – some, not all) have their limits. I would just mention one: their comparative aspect is underdeveloped. Usually, one knows his own society quite well, but rather little about others (and even this is rather sketchy and sometimes severly biased for various reasons). Correspondingly, anthropologists have had more success in describing folk theories than in comparing them cross-culturally and proposing explanatory mechanisms – even the idea of comparing cultures was (is?) epistemiologically-dubious for some anthropologists.

    After such a fairly boring defence of anthropologists, a funny thought came to mind… Why not turn Trivers’ hypothesis on its head and think that maybe self-deception could be a good thing for evolutionary theorists as real human beings? I don’t think that an evolutionary psychologist consciously screens out pub invitations by the genetic fitness of the opposite-sex proposer, compares the genetic relatedness of her kindred with the value and intensity of Christmas cards received, or lays down her life for the life of three siblings or nine cousins (with a profit!). Or, to paraphrase an old economists’ joke, would she not tip a genetically-unrelated waiter she will never meet again? Jokes aside, it would be interesting to know how evolutionary-minded scientists manage to balance their academic knowledge with social performance (and here comes that folk anthropology mentioned above) whenever these two would lead to conflicting commitments!

    Sorry for the long comment, but this nice post made me quite verbose 🙂