50 words for snow 2: or, the economics of cultural categories

Humans evolved in a world where knowing whether an animal was an antelope or a lion was essential for their survival: they could eat the antelope, and they could be eaten by the lion. Accordingly, the human mind seems to have evolved to organize its knowledge of the natural world into sets of related categories (e.g., cheetahs and lions are kinds of felines, aardwolves and spotted hyenas are kinds of hyenas, and felines and hyenas are kinds of carnivores) and to use these relations to make inductions about those categories (e.g., if lions digest fast, then aardwolves probably digest fast, too). That is, humans appear to be innately prepared to build taxonomies which guide their inductions about the world. [López, et al 1997, 252]

It may be asked (and has been by Jeb in the comments to the last post) how it is that categories are subverted or made unnatural by culture. I think this is a misconstrued of the issue. It is not that categories, at least at the ethnotaxonomy level, somehow get things wrong. Instead it is better to ask what it is that categories track. Categories are generally stable ways to divide the regularities of an environment. In short, they model those regularities in ways that lessen our cognitive load, for essentially practical reasons. That is, a category is something we use to do things that matter.

But the environment of any cognitive community includes, amid the natural world, the social world and its techniques, rituals and expectations of its members. These things are regularities in the environment, and knowing them is as important as knowing the general categories about other organisms and conditions than the human. Indeed, one of the explanations for enhanced human cognition, the so-called Machiavellian Hypothesis, supposes that one of the most complex categorical domains humans adapted to dealing with is other humans. Social categories are as important as knowing what predators do, and for much the same reason.

It is common for critics and advocates both of cognitive adaptation theories to say something akin to what López et al. say above: that cognitive adaptation is about survival. But this overlooks a major point: that evolution through natural selection (in proper terms, adaptation by natural selection) is not about survival as such. It is about long term multigenerational success at propagating biological and cultural dispositions. This needs some unpacking.

Consider two variations in a population. One generates many progeny per generation; the other only a few. Naively, we might expect the former to swamp the population with its genes, and hence be fitter. But suppose the cost of many progeny is less-week-fed progeny in times of scarcity, while the fewer progeny share in more food per capita. As a result, the fecund variant has progeny that are less robust in times of harshness. If the environment frequently gets harsh, say, every winter, all or most of the fecund variant might die before getting to reproductive age, while the robust less-fecund survive to reproduce. Or it might be the fecund less-robust progeny do not bring their own progeny to term as frequently as the less-fecund variant. Either way, after a few generations (arbitrarily, geneticists use the F3 generation as the marker), the less-fecund variant will dominate the population. Fitness is defined as long term success in an environment.

Now consider a cultural variant – categorising social behaviours, etc., more nearly like those of the rest of the population/culture. If being in tune with your community increases the likelihood that you will do well* (get work, help when needed, protection for your family, etc.), then that category is also cognitively adaptive. So any categorical disposition towards coadaptation of categories is likely to spread also. I say “disposition” as it is clear that concepts and cultural practices are not predetermined at birth by geographical population, contrary to some claims. A category must be dispositionally agreeable, or it costs too much to acquire (unless the benefits of acquiring it outweigh the costs, as in a community of mathematicians).

Thus, we might expect that a category persists in a culture for either being a reflection of the asocial environment, or being a reflection of the social environment, or both. This explains the Karam view of the cassowary. The Karam see cassowaries as human because of cultural norms, defeating the natural categories of the biotic environment. But why do they see cassowaries as not birds in the first place? It is one thing to explain the persistence of that categorisation. It is another to explain why that occurred in the first place. And here we must appeal to two explanatory principles: costly signalling, and contingency.

Costly signalling is a theory from evolutionary biology in which apparently maladaptive traits evolve as ways to honestly signal fitness. The classical examples are the peacock’s tail, and the “stotting” behaviour of gazelles. Peacocks carry a large tail that impedes flight but signals to potential mates they are able to overcome this liability and thus are fit mates. These types of exaggerated traits evolve due to the fact that some feature is a mating cue and thus it gets more and more selected for by the choosey sex until it reaches a tradeoff point between being a sexual cue and being too onerous a liability. The stotting behaviour, on the other hand, is where gazelles that are being hunted leap high into the air to signal to the predator they are fit enough to not be worth the effort of chasing. In both cases, there is a cost to the signal which makes it an honest one.

In cultural evolution, costly signals honestly indicate that the one doing the signalling is committed to the ethny (the cultural group**), as the signals incur a cost that exceeds what a free-rider would be prepared to pay to get the benefits of that ethny. Thus, avoiding a food source that might be useful in hard times, like pork, indicates commitment to the group. Likewise, spending time in youth acquiring ritual knowledge, accents, and the sacred and cultural items needed to be a part of a community, all of which come at a cost, honestly signal commitment. This is used especially to explain religious commitment (Bulbulia 2004), but it also applies to “racial” groups, political affiliations, and, here, intellectual schools or traditions.

The Karam forego the use of cassowaries as resources in part because to do so is a (mildly) costly signal of group commitment. I have previously argued that this is why religions believe in silly notions (here, herehere, and here). But this fails to explain two things on its own. One is that it fails to explain why relatively low cost miscategorisations play that role (if cassowaries have little use in that ethny anyway). It also fails to explain why that category is the one used as an honest signal. And to do this I think we need to appeal to historical contingency; that is, accidents. There is little explanatory power in accidental events. However, if we want to explain why the QWERTY keyboard is used despite its obvious shortcomings, we must appeal to the fact that the QWERTY keyboard was the one used by the first really successful typewriter, and the costs involved in learning and training users of keyboards outweighs the benefits of adopting a better structured keyboard like the Dvorak layout. QWERTY is just a frozen historical accident.

Screenshot 2016 06 02 10 45 45

The honest cost here is not in the adoption of the convention, but in the replacement of it. For a Karam to change their view that the cassowary is not a bird is to lose all of the conceptual, cultural and ritual behaviours that rely upon that ethnotaxonomy, rather like a creationist replacing their belief that the earth is 6000 years old and static leads to a loss of community acceptance, moral principles, and expectations.

How does this tie into scientific categorisations? One might think of the way in which Freudian categories for mental illness found their way into the DSM (Murphy 2006). Despite repeated research showing that the categories are not “natural” (such as “borderline personality disorder, which was on the boundaries between the Freudian categories “psychosis” and “neurosis”, both of which have been massively revised), those categories are tied into treatment and funding options and general psychiatric institutionalised behaviours to the point that simply abandoning them is too costly in the community of psychiatry.

Another instance is one in which I have a hunting dog: taxonomy ranks in biology. The ranks of both species and genus remained strongly in use despite considerable evidence and argument that neither represents any natural facts about the world, simply because the naming and mentions of genera and species are now tied into both the extensive descriptive literature and educational schemes, as well as such public legal practices like conservation law and programs.

Science is a community of researchers, and like all communities, requires signals of commitment. However, also like all communities, historical accidents and costly signals play a role in the ways in which categories are adopted. In said in the last post that science is adapted to eliminating miscategorisations; but is it really? When it fails, why does it fail? When it succeeds, why does it succeed? I will give my arguments about this in the next post. But for now it pays to conclude that all such classifications represent something regular in the environment. The question to be answered is, what aspects of the environment? Social or not? If all scientific categories are at least partly constructed, is science merely conventional in its categorisations?

Stay tuned…


  • That is, it’s not a psychotic culture with rules like “Kill those who think like we do”. But then, such a society is unlikely to persist long as a society.

** “An ethny can be represented … as a cluster of overlapping, egocentered, concentric kin circles, encompassed within an ethnic boundary.” Van den Berghe 1981, page 22. The term is roughly identical to “ethnic group”, but I use it to denote any cultural cluster that individuals identify with.


Bulbulia, J. (2004). Religious costs as adaptations that signal altruistic intention. Evolution and Cognition 10(1): 19–42.

López, A., et al. (1997). The tree of life: Universal and cultural features of folkbiological taxonomies and inductions. Cognitive Psychology 32: 251–295.

Murphy, D. (2006). Psychiatry in the scientific image. Cambridge, MA; London, MIT Press.

Van den Berghe, P. L. (1981). The ethnic phenomenon. New York; London, Elsevier

2 thoughts on “50 words for snow 2: or, the economics of cultural categories

  1. That culturally mandated ideas and practices serve as the basis for social solidarity and are thus globally adaptive is reasonable enough, but just calling them frozen accidents like QWERTY leaves an awful lot unexplained. In any case, most segments of human culture don’t look much like random choices. They have an apparent internal coherence, just not one obviously based on adaptiveness. I guess you could invoke the notion of spandrels here to fill out the intricate details of human civilization—on this account, the spandrels of San Marco were, in fact, spandrels. But are most of the features of our world similarly spandrels? If so, perhaps it makes sense to think of adaptation as putting a constraint on the space of possible civilizations rather than expecting it to explain very much.

    When it’s a question of somatic, genetically programmed adaptations like muscle attachments or bone shapes, you can use biochemistry and physics to explain how the arrangements produce greater fitness or perhaps are just side effects of what really gets the job done. Rather more effort is required to make sense of the particulars of Christian theology or French verbs or, for that matter, totemism in Papua. I suspect that a lot of what goes on is what the Russian formalists called the motivation of the device. Once generative systems like languages, myths, or sciences are in place, the internal logic of the systems tends to take over, at least in the short run. You don’t argue legal cases by appealing to selection pressures in the Paleolithic.

    A lot more is in play, obviously. For example, whether or not a no-nonsense, cash value pragmatism is the most adaptive philosophy for a society to adopt is an empirical question. We might be better off thinking that cassowaries aren’t birds, i.e., privileging the traditions of the tribe over the facts isn’t obviously an inferior strategy from the point of view of inclusive fitness. What it is for an individual circe 2017 is a statement of ethical preference. A Marxist might say it is a manifestation of the relative power of material over cultural or social capital. Even if, counterfactually, we somehow knew what epistemological stance would maximize our inclusive fitness, it would still be a choice to give a damn about inclusive fitness. There are, after all, other and arguably better things to care about. Or so this heretic asserts.

    I always feel like I should apologize for thinking out loud in these responses, John. Your post was impressively thoughtful. I’m afraid I’m just offering a few fragmentary reservations.


  2. “how it is that categories are subverted or made unnatural by culture.”

    I think I would frame it as how do they become natural and stable over- time.

    ‘frozen historical accident’

    Its certainly a common observation in regard to the conservative nature of folklore. Origin + static repetition.

    I think it has to be more dynamic. As the context for the telling dies you have to find another to hold words on the breath and keep them in memory. An oral culture needs a space in which things are brought to mind and we become compelled to speak, see and hear them.

    It suggests such things will be used in multiple contexts and when the cultural alters they are able to alter as well and hold form in a different space and environment.

    Not frozen, constant state of movement that never stops. Things that in the next moment can be found elsewhere.


Leave a Reply