Report from the Consilience conference, part III

by Massimo Pigliucci

[This is a report from the consilience conference held at the University of Missouri-St. Louis. Part I is here, part II here]

Last day of the consilience conference! We began with David Sloan Wilson, with whom I had just had a spirited (and constructive) discussion about how to measure individual and group cultural selection quantitatively (he admitted it hasn’t been done, yet...). His topic was “Using evolution to improve the quality of life.” According to David, evolutionary principles can be used to improve our quality of life at the level of cities and neighborhoods. His Evolutionary Institute is a think tank that has the explicit goal of connecting evolutionary ideas to public policy.

Wilson sees symbolic thought as an inheritance system, which is necessary to get a theory of cultural evolution off the ground if one excludes fuzzy concepts like memes (which he apparently is inclined to do). He proposes the idea of a “symbotype”-phenotype relationship similar to the standard genotype-phenotype relationship in evolutionary biology. (Though it has to be noted that genotype-phenotype mapping is one of the most difficult problems facing evolutionary biologists, and I doubt that it’s going to be any easier to operationalize the concept of symbotype-phenotype mapping.)

David’s example was a study of prosociality in Binghamton, NY neighborhoods. He geo-mapped individuals who had been scored on a measure of prosociality. He then ran statistical analyses exploring the social correlatives of prosociality on the territory. Prosociality turned out to vary over very small spatial scales. The survey found a strong correlation between the prosociality of individuals and that of their social environment (i.e., the more socially supportive the environment is, the more prosocial people are). Multiple forms of social support contribute to individual prosociality, and adolescents changing location within the city is taken by Wilson to demonstrate phenotypic plasticity. (I actually worked on phenotypic plasticity as it is understood in biology, and I think this is a somewhat metaphorical use of the term.)

This was all very interesting, and even useful from a practical (policy) perspective. But it is social science, the results are unlikely to surprise a social scientist, and the e-word did not add anything at all, as far as I could see, to the whole picture. Evolution has to do with fitness-related variation and inheritance. There were no measures of fitness in the data, and it’s hard to see in what sense an individual changing behavior from one year to another (e.g., moving to a different neighborhood) counts as “inheritance.” But maybe I missed something crucial, somewhere.

Next we moved to Barbara Oakley, on cold-blooded kindness: insights into pathological altruism. This is a situation where while the underlying motivation is to help others, the altruistic behavior has irrational and substantially negative consequences to the other and to the self. An example presented by Oakley was of a woman who married a drug addict and then killed him in self-defense. (More on this below.)

Oakley comes to this as an engineer, and she seeks analogies between engineering and social science principles. For instance, she takes the idea that local decreases in entropy must be offset by a broader increase in entropy in the environment (which is a fundamental principle of physics) and translates it into the idea that even good deeds can carry negative consequences in the human realm. This, honestly, seems to be a stretch, and not a novel insight, given that both social scientists and philosophers have explored the idea of consequentialism in detail, and without needing to reference entropy.

(I’m beginning to think that it would have been nice to have actual social scientists, historians and assorted humanists at this conference, just to see how they would have reacted to biology-based criticism of their fields. Another time, maybe.)

Back to the story of the woman who killed her husband. Apparently, it wasn’t self-defense at all, it was premeditated (he was shot in the back, and she had pre-dug his grave). She was also a sadomasochist, who had complained in the past that her husband refused to hurt her. Oakley contends that the fact that reporters for the National Inquirer (where she originally heard of the story) and others felt sympathy for the woman and bought into her side of the story (though apparently neither the prosecutor nor the jury did) is because we are at fault for excessive (pathological) altruism. Again, that seems a stretch. First, if we were given the actual facts, instead of the National Inquirer version, I bet very few people would have felt compassion for the woman. Second, this case needs to be understood against a cultural background — which Oakley did refer to — of a number of stories of battered women who do act, truly, in self-defense. None of the above, as interesting as it was, had much to do with consilience, as far as I could tell.

Next: Jonathan Gottschall on the storytelling animal, how stories make us human. We all know that we like fiction and stories, but we are not aware of just how much. Storytelling is about someone else in a sense taking over emotional control of your self for a while. We don’t leave storytelling when we go to sleep: dreams, whatever actual physiological function they have, are fragments of stories which the brain tells itself. And then there is daydreaming, in which we apparently spend a large chunk of our day (this includes, however, rethinking past actions and situations, or imagining how to handle likely future actions and situations).

The left hemisphere of the brain is known to be a storyteller, in charge of making sense of everything that happens to us. The downside is that when it doesn’t have reliable information the left “interpreter” will make up stories anyway. Classical experiments show that even simple moving geometrical shapes on a screen are interpreted by many people as agents interacting with each other, driven my motives. (To be fair, since Gottschall showed an example on screen, the shapes were moving around the screen in obviously non random fashion. Would people make up stories if the movements were random?)

Fictional stories have surprisingly wide ranging effects, for instance in the case of the (alleged) role of Uncle Tom’s Cabin in the events that led to the American Civil War. Or D.W. Griffith’s The Birth of a Nation, credited with (temporarily) reviving the then moribund Ku Klux Klan.

Why storytelling? It could be the result of sexual selection, or a byproduct of the way the human mind works, and there are other possibilities. Gottschall seems to favor a pluralist answer, with multiple causes for the origin of story telling propensities in humans. The obvious question is how one would go about testing these hypotheses.

Fiction has a “universal grammar”: character + predicament + attempted solution. Right, and not really surprising. Of course this tells you close to nothing about individual stories and how they vary with culture and time, but point taken. If story telling has a function, it may be a sort of “flight simulator” of the mind, through which we practice for life (do we need to practice possible encounters with zombies?). There is some preliminary evidence, apparently, that people who enter into fictional simulation more often do better in real life.

The first speaker of the afternoon was Henry Harpending, on kinship within populations: whole genomes as green beards. The green beards refer to Dawkins’ (hypothetical) example of people with green beards being inclined to help others sporting the same trait, a process undermined by how easy it is to fake a green beard. This is obviously relevant to the idea of kin selection-mediated altruism and how it is vulnerable to cheaters. Harpending asked how much evidence do we have for mechanisms of kin recognition (to counter cheaters) having evolved in humans. Not much, apparently.

Harpending went on to compare two versions, from research in the ‘70s, of “Mr. Natural”: on the one hand the cooperative and peaceful bushmen of the Kalahari desert; on the other hand the Yanomamo of the Amazons, fierce and aggressive. The question, naturally, is how can these two so different cultures both represent “Mr. Natural”? The most recent take is that there is no such thing as Mr. Natural, that people’s ways of living change rapidly from time to time and culture to culture.

After that excursion, Harpending returned to kinship, and how these days we can actually measure it via Single Nucleotide Polymorphisms (SNPs), an increasing large database which is becoming available for humans through projects like 23andme (a commercial genome sequencing enterprise). The result seems to be that there is not enough dispersion of kinship values in, say, the modern French or Japanese populations, to make it effective to deploy cryptic kinship detection mechanisms. However, this does not hold for small human populations, like the inhabitants of Surui island. In that case there is a significant spread of kinship coefficients.

Again, while much of this was interesting, it wasn’t at all clear to me what it had to do, if anything, with the theme of the conference. It isn’t unusual for people to be invited to a conference with a central theme, show up and then talk about whatever struck their fancy most recently. But there seemed to be a particularly high occurrence of this at the consilience conference.

The next speaker was Pascal Boyer, on “the dark matter of human history, the perils of cognition blindness.” Social science “that matters” needs to address questions like why people engage in warfare, why is there religion, and so on. Boyer is explicitly using the term consilience as synonymous with integration of disciplines, which is, again, different from E.O. Wilson’s use of the term.

Boyer uses an analogy with dark matter in physics to signify that there is quite a bit in social science that does not meet the eye, and that has to do with the neuro-cognitive processes underlying conscious mental states, motivations, social interactions etc. Parts of human nature are not accessible to conscious inspection and cognition awareness requires effort. This notwithstanding, intuitive (or “folk”) sociology takes an intentional stance to groups and states (memories, beliefs), so that behavior can be interpreted as goal-directed.

For Boyer intuitive sociology may be adaptive in the context of our social evolution, as we can trace what happens around us as the result of intentions by other people. However, intuitive sociology fails when it is applied to, for instance, understanding the economy (“folk economics”). An example, apparently, is the idea of rent controls. They make sense from the point of view of folks economics, because the landlord and the lodger are given intentional stances; they don’t make sense for economists because offers depend on preferences and availability of means in the relevant population. (This seems to entirely ignore the fact that rent control measures are usually implemented not to solve an economic problem, but to minimize negative social side effects of purely economic “solutions.”)

Boyer also criticized political science for having empirical basis but no theory, resulting in either formal modeling that does not make contact with empirical reality or the study of political preferences as brute facts.

After a long bit on warfare and ethnic conflict, where he stressed the same contrast between “folk” and more sophisticated theories of what goes on, Boyer concluded by stating that there is no such thing as religion. We expect of religions that they have doctrines, beliefs, personnels (priests, shamans, etc.), and domains of competence, such as survival after death, morality, etc.. But in reality, in most cases at the tribal level — argues Boyer — there is no doctrine at all, the personnel is varied and ad hoc, and there is no unified domain of religion. Seems to me that here as in other examples during the talk there is a confusion between origins on one hand and development and maintenance on the other hand of a given phenomenon. Religions may have originated without the characteristics of the modern stuff, but this neither licenses the bold claim that religion “doesn’t exist,” nor does it imply that things like modern religions doctrines, beliefs, personnel etc. don’t need to be understood on their own terms. Boyer is correct, however, in separating the issue of religion as a type of social organization that is typical of some human societies from supernaturalism, which seems to be a human universal. A consequence he derives is that it makes no sense to think of religion as an adaptation, as it is far too much of a late comer in human evolution.

The take home message is that the social sciences have been disappointing because they have not addressed the big questions, leading to no cumulative progress. (The latter, I think, is a bit uncharitable.) Things went wrong because of the lack of vertical integration, in this case a reduction to neuroscience. Again, Boyer seems to be making a couple of common mistakes at this conference. First, integration and reduction are different things. Second, reduction does not eliminate the higher level phenomena, it only helps explaining them. So, I think, social science should still focus on the high level targets, but also integrate as much as possible notions from other disciplines, including but not limited to neuroscience.

And last: Patricia Churchland on how the mind makes morals. Darwin (together with Hume and Aristotle) thought that our moral sense is the result of social instincts, habits and reason. Churchland’s basic hypothesis is that sociability is of value for social mammals and it evolved by natural selection; its neural “hub” is the hormone oxytocin (involved in attachment, bonding); this is augmented by the reward system in the brain; and the whole thing is elaborated by prefrontal structures in the brain.

Attachment and trust are the platform for moral values. Social problem solving is mediated by the enlarged prefrontal cortex, which overrides, represses, calculates and plans. Refreshingly, Churchland isn’t trying to “reduce” culture to neurobiology, she is after the much more sensible goal of understanding the neural structures that make it possible for us to have culture to begin with.

She cites Eleanor Rosch’s work on the “radial” structure of concepts, with a prototype at the center and fuzzy boundaries. (This is similar to family resemblance in Wittgenstein, which a philosopher like Churchland should have noted.) Social categories are also radial, including the category of “moral.” Interestingly, artificial neural networks “learn” to categorize by way of prototypical structures and fuzzy boundaries. The idea, of course, is that the brain is relevantly similar to neural networks, and likely learns in a similar fashion (which is interesting, but let’s not forget that the brain — unlike neural networks — comes with a great deal of genetic-developmental prewiring).

One final comment on the entire conference: I got the impression that a number of participants did not actually read Wilson’s Consilience, at least not recently (several have admitted as much to me). Many (though not all) seemed to be convinced that the book promotes a positive and multi-directional exchange between disciplines, particularly crossing the science-humanities divide. It does nothing of the sort. It is a clear attempt at a reductionist program of subsuming the humanities into the sciences, and particularly biology, though it isn’t obvious why Wilson didn’t go all the way and subsume biology itself into physics. Perhaps because he’s a biologist?

—
Interesting footnote: one of the conference attendees heard that I was blogging about it, and objected to it, on two grounds: first, I am bringing to an outside forum discussions and opinions that were not meant for that forum; second, I get to editorialize and comment about what was said at the conference, while the other participants can’t.

My response is that conferences of this type are public events (registration was open to everyone), and that bringing at least a flavor of the proceedings to a wider audience is a good thing (at least another participant was Tweeting about it, by the way). As for commenting, well naturally I am writing this, so you are getting my take on it. Presumably, the intelligent reader is aware of this and will take it into account in forming her own judgment. Moreover, once my thoughts are out in the blogosphere other participants can either comment on them directly or can use other forums to respond to and/or expand upon them.

Still, the question does raise interesting issues concerning the ethics of blogging from academic conferences (or in other situations), and I’d be interested in hearing people’s thoughts on this.