A rare example of cultural evolution in action:
Throughout the Highlands of New Guinea, a group’s ability to raise large numbers of pigs is directly related to its economic and social success in competition with other regional groups. The ceremonial exchange of pigs allows groups to forge alliances, re-pay debts, obtain wives, and generate prestige through excessive displays of generosity. All this means that groups who are better able to raise pigs can expand more rapidly in numbers—by reproduction and in-migration—and thus have the potential to expand their territory. Group size is very important in intergroup warfare in small-scale societies so larger groups are more likely to successfully expand their territory. However, the prestige more successful groups obtain may cause the rapid diffusion of the very institutions, beliefs, or practices responsible for their competitive edge as other groups adopt their strategies and beliefs.
In 1971, the anthropologist David Boyd was living in the New Guinea village of Irakia, and observed intergroup competition via prestige-biased group transmission. Concerned about their low prestige and weak pig production, the senior men of Irakia convened a series of meetings to determine how to improve their situation. Numerous suggestions were proposed for raising their pig production but after a long process of consensus building the senior men of the village decided to follow a suggestion made by a prestigious clan-leader who proposed that they “must follow the Fore’” and adopt their pig-related husbandry practices, rituals, and other institutions. The Fore’ were a large and successful ethnic group in the region, who were renowned for their pig production. The following practices, beliefs, rules, and goals were copied from the Fore’, and announced at the next general meeting of the community:
1) All villagers must sing, dance and play flutes for their pigs. This ritual causes the pigs to grow faster and bigger. At feasts, the pigs should be fed first from the oven. People are fed second.
2) Pigs should not be killed for breaking into another’s garden. The pig’s owner must assist the owner of the garden in repairing the fence. Disputes will be resolved following the dispute resolution procedure used among the Fore’.
3) Sending pigs to other villages is tabooed, except for the official festival feast.
4) Women should take better care of the pigs, and feed them more food. To find extra time for this, women should spend less time gossiping.
5) Men must plant more sweet potatoes for the women to feed to the pigs, and should not depart for wage labor in distant towns until the pigs have grown to a certain size.
The first two items were implemented immediately at a ritual feast. David stayed in the village long enough to verify that the villagers did adopt the other practices, and that their pig production did increase in the short term, though unfortunately we don’t know what happened in the long-run.
Let me highlight three features of this case. First, the real causal linkages between many of these elements and pig production are unclear. Maybe singing does cause pigs to grow faster, but it’s not obvious and no one tried to ascertain this fact, via experimentation for example. Second, the village leadership chose to rely on copying institutions from other groups, and not on designing their own institutions from scratch. This is smart, since we humans are horrible at designing institutions from scratch. And third, this transmission between groups occurred rapidly because Irakia already had a political institution in the village, involving a council of the senior members of each clan, who were empowered by tradition (social norms) to make community-level decisions. Lacking this decision-making institution, Fore’ practices would have had to spread among households, and thus been much slower in spreading. Of course, such political decision-making institutions themselves are favored by intergroup competition.
This is it. This is the five-point platform that the Democratic Party can use to win in 2020.
Yesterday’s review mentioned that children have certain “slots” in their heads that are ready for specific types of incoming information. How far can we take this idea?
The UCLA anthropologist Dan Fessler argues that during middle childhood (ages 6-9) humans go through a phase in which we are strongly attracted to leraning about fire, by both observing others and manipulating it ourselves. In small-scale societies, where children are free to engage this curiosity, adolescents have both mastered fire and lost any further attraction to it. Interestingly, Fessler also argues that modern societies are unusual because so many children never get to satisfy their curiosity, so their fascination with fire stretches into the teen years and early adulthood.
On prestige-based socialization and learning who to learn from:
By 14 months, infants are already well beyond social referencing and already showing signs of using skill or competence cues to select models. After observing an adult model acting confused by shoes and placing them on his hands, German infants tended not to copy his unusual way of turning on a novel lighting device: using his head. However, if the model acted competently, confidently putting shoes on his feet, babies tended to copy the model and used their heads to activate the novel lighting device.
Kind of unrelated to culture, but did you know quadruped animals run at quantized speeds?
Many four-legged animals are saddled with a design disadvantage. Game animals thermoregulate by panting, like a dog. If they need to release more heat, they pant faster. This works fine unless they are running. When they run, the impact of their forelimbs compresses their chest cavities in a manner that makes breathing during compressions inefficient. This means that, ignoring oxygen and thermoregulation requirements, running quadrupeds should breathe only once per locomotor-cycle. But, since the need for oxygen goes up linearly with speed, they will be breathing too frequently at some speeds and not frequently enough at other speeds. Consequently, a running quadruped must pick a speed that (1) demands only one breath per cycle, but (2) supplies enough oxygen for his muscle-speed demands (lest fatigue set in), and (3) delivers enough panting to prevent a meltdown (heat stroke), which depends on factors unrelated to speed such as the temperature and breeze. The outcome of these constraints is that quadrupeds have a discrete set of optimal or preferred speed settings (like the gears on a stick-shift car) for different styles of locomotion (e.g., walking, trotting and galloping). If they deviate from these preferred settings, they are operating less efficiently.
Humans lack these restrictions because (1) our lungs do not compress when we stride (we’re bipedal) so (2) our breathing rates can vary independent of our speed, and (3) our thermoregulation is managed by our fancy sweating-system, so the need to pant does not constrain our breathing. Because of this, within our range of aerobic running speeds (not sprinting), energy use doesn’t vary too much. That means we can change speeds within this range without paying much of a penalty. As a result, a skilled endurance hunter can strategically vary his speed in order to force his prey to run inefficiently. If his prey picks an initial speed just faster than the hunter, to escape, the hunter can speed up. This forces the prey to ‘shiftup’ to a much faster speed, which will cause rapid overheating. The animal’s only alternative is to run inefficiently, at a slower speed which will exhaust his muscles more quickly. The consequence is that hunters force their prey into a series of sprints and rests that eventually result in heat stroke. The overheated prey collapses, and is easily dispatched. Tarahumara, Paiute and Navajo hunters report that they then simply strangle the collapsed deer or pronghorn antelope.
Even locomotion is culturally learned!
To achieve a running form that maximizes both performance and freedom from injury, humans need to rely on some cultural learning, on top of much individual practice. The evolutionary biologist and anatomist, Dan Lieberman, has studied long-distance barefoot and minimally shod running in communities around the globe. When he asks runners of all ages how they learned to run, they never say they “just knew how.” Instead, they often name or point to an older, highly skilled, and more prestigious member of their group or community, and say they just watch him, and do what he does. We are such a cultural species that we’ve come to rely on learning from others even to figure out how to run in ways that best harness our anatomical adaptations.
Why we use spices:
Why do we use spices in our foods? In thinking about this question keep in mind that (1) other animals don’t spice their foods, (2) most spices contribute little or no nutrition to our diets, and (3) the active ingredients in many spices are actually aversive chemicals, which evolved to keep insects, fungi, bacteria, mammals and other unwanted critters away from the plants that produce them.
Several lines of evidence indicate that spicing may represent a class of cultural adaptations to the problem of food-borne pathogens. Many spices are antimicrobials that can kill pathogens in foods. Globally, common spices are onions, pepper, garlic, cilantro, chili peppers (capsicum) and bay leaves. Here’s the idea: the use of many spices represents a cultural adaptation to the problem of pathogens in food, especially in meat. This challenge would have been most important before refrigerators came on the scene. To examine this, two biologists, Jennifer Billing and Paul Sherman, collected 4578 recipes from traditional cookbooks from populations around the world. They found three distinct patterns.
1. Spices are, in fact, antimicrobial. The most common spices in the world are also the most effective against bacteria. Some spices are also fungicides. Combinations of spices have synergistic effects, which may explain why ingredients like “chili power” (a mix of red pepper, onion, paprika, garlic, cumin and oregano) are so important. And, ingredients like lemon and lime, which are not on their own potent anti-microbials, appear to catalyze the bacteria killing effects of other spices.
2. People in hotter climates use more spices, and more of the most effective bacteria killers. In India and Indonesia, for example, most recipes used many anti-microbial spices, including onions, garlic, capsicum and coriander. Meanwhile, in Norway, recipes use some black pepper and occasionally a bit of parsley or lemon, but that’s about it.
3. Recipes appear to use spices in ways that increase their effectiveness. Some spices, like onions and garlic, whose killing power is resistant to heating, are deployed in the cooking process. Other spices like cilantro, whose antimicrobial properties might be damaged by heating, are added fresh in recipes.
Thus, many recipes and preferences appear to be cultural adaptations adapted to local environments that operate in subtle and nuanced ways not understood by those of us who love spicy foods. Billing and Sherman speculate that these evolved culturally, as healthier, more fertile and more successful families were preferentially imitated by less successful ones. This is quite plausible given what we know about our species’ evolved psychology for cultural learning, including specifically cultural learning about foods and plants.
Among spices, chili peppers are an ideal case. Chili peppers were the primary spice of New World cuisines, prior to the arrival of Europeans, and are now routinely consumed by about a quarter of all adults, globally. Chili peppers have evolved chemical defenses, based on capsaicin, that make them aversive to mammals and rodents but desirable to birds. In mammals, capsicum directly activates a pain channel (TrpV1), which creates a burning sensation in response to various specific stimuli, including acid, high temperatures and allyl isothiocyanate (which is found in mustard or wasabi). These chemical weapons aid chili pepper plants in their survival and reproduction, as birds provide a better dispersal system for the plants’ seeds than other options (like mammals). Consequently, chilies are innately aversive to non-human primates, babies and many human adults. Capsaicin is so innately aversive that nursing mothers are advised to avoid chili peppers, lest their infants reject their breast (milk), and some societies even put capsicum on mom’s breasts to initiate weaning. Yet, adults who live in hot climates regularly incorporate chilies into their recipes. And, those who grow up among people who enjoy eating chili peppers not only eat chilies but love eating them. How do we come to like the experience of burning and sweating—the activation of pain channel TrpV1?
Research by psychologist Paul Rozin shows that people come to enjoy the experience of eating chili peppers mostly by re-interpreting the pain signals caused by capsicum as pleasure or excitement. Based on work in the highlands of Mexico, children acquire this gradually without being pressured or compelled. They want to learn to like chili peppers, to be like those they admire. This fits with what we’ve already seen: children readily acquire food preferences from older peers. In Chapter 14, we further examine how cultural learning can alter our bodies’ physiological response to pain, and specifically to electric shocks. The bottom line is that culture can overpower our innate mammalian aversions, when necessary and without us knowing it.
Fascinating if true. But don’t we use spices because of their taste? If spices are antimicrobials, why aren’t there any tasteless spices? I guess you could argue most plants taste like something, usually something bad, and if a plant is a good antimicrobial then we go through the trouble of culturally reinterpreting its taste to be “exciting” or “interesting”. Also, how far can this “cultural reinterpretation” idea go? Does this explain things like masochism, or like the weak form of masochism that makes people like naively unpleasant experiences like roller coasters?
I knew that Europeans had light skin because they lived in northern latitudes without much sunlight. But then how come Inuit and North Asians never developed light skin? Henrich explains:
To understand this, we need first to consider how culture has shaped genes for skin color over the last 10 millennia. Much evidence now indicates that the shades of skin color found among different populations—from dark to light—across the globe represent a genetic adaptation to the intensity and frequency of exposure to ultraviolet light, including both UVA and UVB. Near the equator, where the sun is intense year round, natural selection favors darker skin, as seen in populations near the equator in Africa, New Guinea and Australia. This is because both UVA and UVB light can dismantle the folate present in our skin, if not impeded or blocked by melanin. Folate is crucial during pregnancy, and inadequate levels can result in severe birth defects like spina bifida. This is why pregnant women are told by their physicians to take folic acid. In men, folate is important in sperm production. Preventing the loss of this reproductively valuable folate means adding protective melanin to our epidermis, which has the side effect of darkening our skin.
The threat from intense UV light to our folate diminishes for populations farther from the equator. However, a new problem pops up, as darker skinned people face a potential vitamin D deficiency. Our bodies use UVB light to synthesize vitamin D. At higher latitudes, the protective melanin in dark skin can block too much of the UVB light, and thereby inhibit the synthesis of vitamin D. This vitamin is important for the proper functioning of the brain, heart, pancreas and immune system. If a person’s diet lacks other significant sources of this vitamin, then having dark skin and living at high latitudes increases one’s chances of experiencing a whole range of health problems, including most notably rickets. A terrible condition especially in children, rickets causes muscle weakness, bone and skeletal deformities, bone fractures and muscle spasms. Thus, living at high latitude will often favor genes for lighter skin. Not surprising for a cultural species, many high latitude populations of hunter-gatherers (above 50-55q latitude), such as the Inuit, culturally evolved adaptive diets based on fish and marine animals, so the selection pressures on genes to reduce the melanin in their skin were not as potent as they would have been in populations lacking such resources. If these resources were to disappear from the diet of such northern populations, selection for light skin would intensify dramatically.
Among regions of the globe above 50-55q latitude (e.g. much of Canada), the area around the Baltic Sea was almost unique in its ability to support early agriculture. Starting around 6,000 years ago, a cultural package of cereal crops and agricultural know-how gradually spread from the south, and was adapted to the Baltic ecology. Eventually, people became primarily dependent on farmed foods, and lacked access to the fish and other vitamin D-rich food sources that local hunter-gatherer populations had long enjoyed. However, being at particularly high latitudes, natural selection kicked in to favor genes for really light skin, so as to maximize whatever vitamin-D could be synthesized using UVB light.
Secret Of Our Success spends a lot of time talking about gene-culture coevolution and how we should expect people from different cultures to have different genes. When asked whether this is potentially racist, it argues it’s really maximally anti-racist, because “racism” means “believing in exactly the same racial categories as 19th century racists”, and gene-culture coevolution proves that variation is actually much more widespread than that, so there.
In case you needed proof that high status increases your inclusive fitness:
Chris asked a sample of Tsimane to rank the men in two villages along a number of dimensions, including their fighting ability, generosity, respect, community persuasiveness, ability to get their way, and their number of allies. Each Tsimane’ man could then be assigned a score based on the aggregate results from his fellow villagers. Chris argues that his measures of fighting ability and community persuasiveness provide the best proxies for dominance and prestige, respectively, in this context. He then shows that both of these proxies for social status are associated with having more babies with one’s wife, having more extra-marital affairs, and being more likely to remarry after a divorce, even after statistically removing the effects of age, kin group size, economic productivity and several other factors. Beyond this, the children of prestigious men die less frequently and prestigious men are more likely to marry at younger ages (neither of these effects hold for dominant men). All this suggests that, at least in this small scale society, being recognized as either dominant or prestigious has a positive influence on one’s total reproductive output (children) or mating success over and above the consequences that might accrue from factors associated with status like economic productivity or hunting skills. Not surprisingly, both dominant and prestigious men tended to get their way at group meetings, but only prestigious men were respected and generous.
On the Sanhedrin:
Effective institutions often harness or suppress aspects of our status psychology in non-intuitive ways. Take the Great Sanhedrin, the ancient Jewish court and legislature that persisted for centuries at the beginning of the Common Era. When deliberating on a capital case, its 70 judges would each share their views beginning with the youngest and lowest ranking member and then proceed in turn to the “wisest” and most respected member. This is an interesting norm because (1) it’s nearly the opposite of how things would go if we let nature take its course, and (2) it helps guarantee that all the judges got to hear the least varnished views of the lower ranking members, since otherwise the views of the lowest status individuals would be tainted by both the persuasive and deferential effects of prestige and dominance. Concerns with dominance may have been further mitigated by (1) a sharing of the directorship of the Sanhedrin by two individuals, who could be removed by a vote of the judges, (2) the similar social class and background of judges, and (3) social norms that suppressed status displays.
I like this idea, but I worry it could backfire. Supposing that even the best of us are at least a little tempted to conform, it risks the youngest and least experienced members setting the tone for the discussion, so that the older and wiser members are tempted to conform with people more foolish than themselves. If the wisest people spoke first, at least we could get their untainted opinions and guarantee that any conformity was at least in favor of the opinion most likely to be correct. Overall it seems like they should have gone with secret ballots. I wonder if anyone’s ever done an experiment comparing wisest-first, youngest-first, and secret-ballot decision-making to see if any have a clear advantage. You could do it with one of those “guess the number of jelly beans in this jar” tasks or something, with participants who did well on a test problem clearly marked as “elders”.
On why societies often dictate naming children after their paternal relatives:
In building a broader kinship network, social norms and practices connect a child more tightly to his or her father’s side of the family, in subtle ways. In contrast to many complex societies, mobile hunter-gatherer populations often emphasize kinship through both mom and dad, and permit new couples much flexibility in where they can live after marriage. However, there’s always that problem of paternity certainty for dad’s entire side. Among Ju/’hoansi, mobile hunter-gatherers in the Kalahari Desert in southern Africa, social norms dictate that a newborn’s father—or, more accurately, the mother’s husband—has the privilege of naming the child. These norms also encourage him to name the child after either his mother or father, depending on the infant’s sex. Ju/’hoansi believe name sharing helps the essence of the paternal grandparents live on, and it consequently bonds both the grandparents and the father’s whole side of the family to the newborn. Relatives of the grandparents often refer to the newborn using the same kinship term they use for his or her older namesake—that is, the grandfather’s daughter will call the newborn baby “father.”
This bias to the father’s side is particularly interesting since Ju/’hoansi kinship relationships are otherwise quite gender egalitarian, emphasizing equally the links to both mom’s and dad’s sides of the family. This biased naming practice may help create that symmetry by evening out the imbalance that paternity uncertainty leaves behind. In many modern societies, where social norms favoring the father’s side have disappeared, the effect of paternity certainty emerges as maternal grandparents, uncles and aunts invest more than the same paternal relatives do. Thus, Ju/’hoansi practices link newborns directly to their father’s parents and simultaneously, via the use of close kin terms like “father” and “sister”, pull all of dad’s relatives closer
I wonder if this can be extended to our own practice of kids (mostly) taking their father’s last name rather than their mother’s.
And Joseph Henrich continues with an anecdote I eventually decided to consider cute:
More broadly, in Ju/’hoansi society, sharing the same name is an important feature of social life, which has many economically important implications. Psychologically, creating namesakes may work in two interlocking ways. First, even among undergraduates and professors, experiments suggest that sharing the same, or even a similar, name increases people’s liking for the other person, their perceptions of similarity and their willingness to help that person. In one study, for example, professors were more likely to fill out a survey and mail it back if the cover letter was signed by someone with a name similar to their own name. The perception of similarity suggests that namesakes may somehow spark our kin psychology, since we already know we use other cues of similarity (appearance) to assess relatedness. Second, even if this same-name trick doesn’t actually spark any change in immediate feelings, it still sets the appropriate social norms—the reputational standards monitored by others—which among the Ju/’hoansi specify all kinds of important things about relationships, ranging from meat sharing priorities to water-hole ownership. Norms related to naming or namesake relationships are common across diverse societies, and many people in small-scale societies intuitively know the power of namesakes, as my Yasawan friends with names like Josefa, Joseteki and Joseses often remind me. My own kids are named Joshua, Jessica and Zoey, thus matching my own first name by first initial or by rhyming.
(his wife is also an anthropologist, so maybe that makes naming your kids according to anthropological phenomena easier to pull off).
Relevant to a frequent discussion here about whether polyamory is “unnatural” or at least a violation of Chesterton’s Fence:
Even in societies with marriage, social norms and beliefs need not re-enforce concerns about sexual fidelity that arise from male pair-bonding psychology, but can instead promote investment in children in other ways. Many South American indigenous populations believe that a child forms in his or her mother’s womb through repeated ejaculations of sperm, a belief system that anthropologists have labeled partible paternity. In fact, people in many of these societies maintain that a single ejaculation cannot sustain a viable pregnancy, and men must “work hard” with repeated ejaculations over many months to sustain a viable fetus. Women, especially after the first fetus appears, are permitted, and sometimes even encouraged, to seek another man, or men, to have sex with in order to provide ‘additional fathers’ for their future child. Anyone who contributes sperm to the fetus is a secondary father. In some of these societies, periodic rituals prescribe extramarital sex after successful hunts, which helps establish and formalize the creation of multiple fathers. Secondary fathers — often named at birth by the mother — are expected to contribute to the welfare of their children (e.g., by delivering meat and fish), although not as much as the primary father, the mother’s husband. Frequently, the secondary father is the husband’s brother.
Obtaining a second father is adaptive, at least sometimes. Detailed studies among both the Bari’ in Venezuela and the Ache’ show that kids with exactly two fathers are more likely to survive past age fifteen than kids with either one father or three or more fathers.
Importantly, social norms cannot just make male sexual jealousy vanish. Men don’t like it when their wives seek sex with other men. However, rather than being supported by their communities in monitoring and punishing their wives for sexual deviations, they are the one’s acting defiantly—violating social norms — if they show or act on their jealousy. Reputational concerns and norms are flipped around here, so now the husband has to control himself. In the eyes of the community, it’s considered a good thing for an expectant mother to provide a secondary father for her child.
Henrich adds that about 85% of human societies have practiced something other than traditionally-understood monogamy.
Suppose somebody in a weird Californian counterculture scene is trying to decide to what degree polyamory is Chesterton’s-Fence-compliant. They might look around their own social network and find that most of the people they know have organically become polyamorous over the past decade or so, and decide it is the local tradition (and therefore it is good). But they could look on a broader scale and see that most people in their civilization over the past few centuries have been monogamous (and therefore polyamory is bad). Or they could look on an even broader scale and see that most people in the world throughout human history have been non-monogamous (and therefore polyamory is potentially good again). I understand other people’s intuition that the “my civilization, past few hundred years” scale seems important, but I’m not sure how you would non-arbitrarily justify choosing that particular scale instead of others. The strongest argument seems to be something like “Wait two generations to see if it builds strong families”, but I could see going either way.
I mentioned aversion to eating insects in the original review, but Henrich suggests some food taboos are easier to acquire than others:
There is reason to suspect that we humans have an innate susceptibility to picking up meat aversions, due to the tendency of dead animals to carry dangerous pathogens. Thus, we humans are primed to acquire meat taboos over other food avoidances
More on taboos. A lot of taboos were of the form “you personally are not allowed to eat this particular meat or else something terrible will happen to you, so you might as well share it with the less fortunate instead”; this looks like a pretty transparent attempt by cultural evolution to build a social safety net. Henrich asks why these taboos persisted in the face of greed:
A good learner will acquire this rule while growing up and never actually violate it (meat is consumed in public), so he’ll never directly experience eating the tabooed part and not having bad luck. Rare cases of taboo violation that, by coincidence, were followed by bad luck or illness will be readily remembered and passed on (psychologists call this “negativity bias”). Meanwhile, cases of violations followed by a long period when nothing bad happens will tend to be missed or forgetten, unless people keep and check accurate records.
Based on my field experience, any skeptic who questions the taboos will be met with vivid descriptions of particular cases in which the taboos were violated and then poor hunting, illnesses, or bad luck ensued.
This is a huge stretch, but I wonder if you could make an argument that evolution favored confirmation bias because it helped prevent people from questioning their cultural rules.
How social norms are maintained:
In research in the villages of Yasawa Island, my team and I have studied how norms are maintained. When someone, for example, repeatedly fails to contribute to village feasts or community labor, or violates food or incest taboos, the person’s reputation suffers. A Yasawan’s reputation is like a shield that protects them from exploitation or harm by others, often from those who harbor old jealousies or past grievances. Violating norms, especially repeatedly, causes this reputational shield to drop, and creates an opening for others to exploit the norm-violator with relative impunity. Norm violators have their property (e.g., plates, matches, tools) stolen and destroyed while they are away fishing or visiting relatives in other villages; or, they have their crops stolen and gardens burned at night. Despite the small size of these communities, the perpetrators of these actions often remain anonymous and get direct benefits in the form of stolen food and tools as well as the advantages of bringing down a competitor or dispensing revenge for past grievances.
Despite their selfish motivations, these actions act to sustain social norms, including cooperative ones, because—crucially—perpetrators can only get away with such actions when they target a norm-violator, a person with his reputational shield down. Were they to do this to someone with a good reputation, the perpetrator would himself become a norm-violator and damage his or her reputation, thereby opening themselves up to gossip, thefts and property damage. This system, which Yasawans themselves can’t explicitly lay out, thereby harnesses past grievances, jealousies and plain old self-interest to sustain social norms, including cooperative norms like contributing to village feasts.282 Thus, individuals who fail to learn the correct local norms, can’t control themselves or repeatedly make mistaken violations are eventually driven from the village, after having been relentlessly targeted for exploitation.
This sounds sort of like the Icelandic legal system in Legal Systems Very Different From Ours, in that the consequence of breaking the law is that the laws cease to protect you. But viewed from a more critical angle, it also sounds like the modern “tradition” of committing (and/or tolerating) hate crimes against people who don’t conform.
Speaking of hate crimes, Henrich (like me) thinks “racism” is not a natural category. He thinks ethnic hostility is much more natural than racial hostility, with the difference being that race is biological and ethnicity is culture. People are naturally friendly towards people of their own culture and skeptical of people from other cultures, which may or may not follow racial lines. He discusses an experiment in which children are asked to view a puppet playing a game incorrectly:
We can see how deeply norms are intertwined with our folk sociology by returning to the experiments with Max the puppet. The child subjects now encounter Max along with Henri. Max speaks native-accented German but Henri speaks French-accented German. Young German children protested much more when Max —their co-ethnic as cued by accent — played the game differently from the model than when Henri did. Co-ethnics are favored because they presumably share similar norms, but that also means they are subject to more monitoring and punishment if they violate those norms. This appears to hold cross-culturally, as people from places as diverse as Mongolia and New Guinea willingly pay a cost to preferentially punish their co-ethnics in experiments like the Ultimatum Game, over their non-co-ethnics, for norm violations.
This approach to how and why we think about tribes and ethnicity has broader implications. First, intergroup competition will tend to favor the spread of any tricks for expanding what members of a group perceive as their tribe. Both religions and nations have culturally evolved to increasingly harness and exploit this piece of our psychology, as they create quasi-tribes. Second, this approach means that the ingroup vs. out-group view taken by psychologists misses a key point: not all groups are equally salient or thought about in the same way. Civil wars, for example, strongly trace to ethnically or religiously marked differences, and not to class, income or political ideology.  This is because our minds are prepared to carve the social world into ethnic groups, but not into classes or ideologies.
Finally, the psychological machinery that underpins how we think about ‘race’ actually evolved to parse ethnicity, not race. You might be confused by this distinction since race and ethnicity are so often mixed up. Ethnic group membership is assigned based on culturally-transmitted markers, like language or dialect. By contrast, racial groups are marked and assigned according to perceived morphological traits, such as color or hair form, which are genetically transmitted. Our folk-sociologcial abilities evolved to pick out ethnic groups, or tribes. However, cues like skin color or hair form can pose as ethnic markers in the modern world because members of different ethnic groups sometimes also share markers like skin color/hair form, and racial cues can automatically and unconsciously ‘trick’ our psychology into thinking that different ethnic groups exist. And, this byproduct can be harnessed and reified by cultural evolution to create linguistically labeled racial categories and racism.
Underlining this point is the fact that racial cues do not have cognitive priority over ethnic cues: when children or adults encounter a situation in which accent or language indicate ‘same ethnicity’ but skin color indicates ‘different race’, the ethno-linguistic markers trump the racial markers. That is, children pick as a friend someone of a different race who speaks their dialect over someone of the same race who speaks a different dialect.  Even weaker cues like dress can sometimes trump racial cues. The tendency of children and adults to preferentially learn and interact with those who share their racial markers (mistaken for ethnic cues) likely contributes to the maintenance of cultural differences between racially marked populations, even in the same neighborhood.
This ties in to my crackpot theory that the number one way to fight racism in the US is to somehow get everyone speaking exactly the same accent.
In one well-studied case among the Gebusi, in New Guinea, my failure to meet my sister exchange obligations would increase the chances that I would, at some future date, be found guilty of witchcraft.
#out of context quotes
Henrich discusses a theory of intrinsic growth pretty similar to the one in my recent singularity post. But he introduces a neat experimental test: Polynesian islands. On larger islands (ie with higher carrying capacities), technological advance is faster:
Islands or island clusters with larger populations and more contact with other islands had both a greater number of different fishing-tool types and more-complex fishing technologies. Figure 12.2 shows the relationship between population size and the number of tool types. People on islands with bigger populations had more tools at their disposal, and those tools tended to be more sophisticated.
Another team, led by the evolutionary anthropologist Mark Collard, found the same kind of strong positive relationship when they examined forty nonindustrialized societies of farmers and herders from around the globe. Once again, larger populations had more-complex technologies and a greater number of different types of tools.
These effects can even be observed in orangutans. While orangutans have little or no cumulative culture, they do possess some social learning abilities that result in local, population-specific traditions. For example, some orangutan groups routinely use leaves to scoop up water from the ground or use sticks to extract seeds from fruit. Data from several orangutan populations show that groups with greater interaction among individuals tend to possess more learned food-obtaining techniques.
The point is, larger and more interconnected populations generate more sophisticated tools, techniques, weapons, and know-how because they have larger collective brains.
Henrich’s model is actually a little more complicated than mine, because it includes a term for forgetting technology (which actually happens pretty often when the group is small enough!) The more technology the group has, the more likely that one or two things slip through the cracks every generation and don’t get passed on to the kids. That means that most primitive societies are in an equilibrium between the rate of generating and the rate of losing technology, whose exact level depends on the population size:
SOme information was lost every generation, because copies are usually worse than the originals. Cumulative cultural evolution has to fight against this force and is best able to do so in larger populations that are highly socially interconnected. The key is most individuals end up imperfect, worse than the models they are learning from. However, some few individuals, whether by luck, fierce practice, or intentional innovation, end up better than their teachers…
One point the book really drove home is how much of the absolute basics of knowledge are cultural inventions. We laugh at primitive tribes who count “one, two, many”, but the idea of counting more specifically than this was a discovery that had to be discovered by someone, and only survived when there was a context that made it useful:
Many of the products of cumulative cultural evolution give us not only ready concepts to apply to somewhat new problems, and concepts to recombine (bows are projectiles + elastically stored energy) but actually give us cognitive tools or mental abilities that we would not otherwise have. Arabic numerals, Roman letters, the Indian zero, the Gregorian calendar, cylindrical projection maps basic color terms, clocks, fractions, and right vs. left are just some of the cognitive tools that have shaped your mind and mine
Alas, this quote is missing some context from the rest of the book showing just how hard these ideas were to develop. Remember that mathematicians spent a while debating whether “zero” was truly a number, that ancient people had what we consider very confusing concepts around color (even the Greeks were weird about this). Remember that the alphabet – breaking words up into their smallest components – arose only after millennia of logographs and syllabaries, and in some areas never arose at all. There’s even some speculation that basic ideas about introspection and emotion were invented pretty late. Or even:
Subordinating conjunctions like “after”, “before”, and “because of” may have evolved only recently, in historical times, and are probably no more a feature of *human* languages than composite bows are a feature of *human* technological repertoires. The tools of subordination seem less well-developed in the earliest versions of Sumerian, Akkadian, Hittite, and Greek. This makes these languages slow, ponderous, and repetitious to read. ..this is not to say that we humans don’t have some souped-up innate abilities for dealing with hierarchical structures, which may also be useful for making tools or understanding social relationships, but merely that the elegant bits of grammar that permit us to fully harness these abilities were built by cultural evolution.
This kind of thing is why Henrich thinks comparing the IQ of young chimps and human toddlers is fair, but comparing older chimps and older humans isn’t. Older humans have all of these deep-level concepts to work with that make solving even abstract puzzles much easier. This is also plausibly related to the Flynn Effect.
A successful communicator is one who can most effectively be understood, given the local social, environmental, or ecological conditions. As young or naïve learners focus on and learn from more successful communicators—who are using more effective communication tools—cumulative cultural evolution will gradually assemble sign or whistled repertoires, over time, in the same way that it hones kayaks, spears, and boomerangs. Given this, there’s no reason to suspect that such cultural evolutionary processes somehow apply only to whistled or gestural sign languages, and not to typical spoken languages. Thus, spoken languages should—under the right circumstances—show some response to the local acoustic environments and to nonlinguistic social norms, just as whistled and sign languages do. While researchers have done little work on such topics, there’s some preliminary evidence.
Spoken languages vary in their sonority. The sonority of our voices decreases as the airflow used for speech is obstructed and is highest for open vowels, like the /a/ and lowest for so-called voiceless stops like the /t/ in tin. Pronounce each of these sounds and note the difference in the constriction of your airflow. Both vowels and consonants vary in sonority, but vowels generally have much higher sonority than consonants. This means that more sonorous languages tend to have more vowels (e.g., Hawaiian), while less sonorous ones pack the consonants together (e.g., Russian). For the same energy and effort, more sonorous speech sounds can be heard at greater distances and over more ambient noise than less sonorous ones.
If languages adapt culturally, then we can predict that in situations in which people do relatively more talking over greater interpersonal distances with more ambient noise and sound dispersion, languages will be more sonorous. Many environmental variables might influence this, but Robert Monroe, John Fought, and their colleagues reasoned that climate, and specifically temperature, might have a big effect. The idea is simple: in warmer climates, people work, play, cook, and relax outdoors. Compared to life indoors, living outside means that communicators more frequently face the challenges of distance, noise and poor acoustics. Their team generated measures of sonority from word lists for dozens of languages and then looked at the relationship between sonority and measures of climatic temperature, like the number of months per year when it’s below 10°C (50°F).
It turns out that if all you know is climatic temperature, then you can account for about one-third of the variation in the sonority of languages. Languages in warmer climates tend to use more vowels than those in colder climates and rely more heavily on the most sonorous vowel, /a/. For consonants, languages in warmer climates rely more heavily on the most sonorant consonants, like /n/, /l/, and /r/. By contrast, languages in colder climates lean more heavily on the least sonorous vowels, as the /i/ in deep.7
This simple idea can have much nuance added to it. For example, not all warm climates are equally conducive to sonorous speech. In regions with dense forest cover, the advantages of high sonority might be less pronounced, or as the anthropologists Mel and Carol Ember have argued, very cold and windy climates may select against linguistic practices that involve opening one’s mouth widely, due to the increased heat loss. To this they added the idea that social norms about sexual restrictiveness might also influence sonority. Adding both of these nuances to the basic climatic temperature analysis, they managed to account for four-fifths of the variation in the sonority of language.
I’m a little worried about p-hacking here, but still, whoa! The thing where Inuit languages sound like tikkakkooktttippik but Polynesian languages sound like waoiuhieeawahiaii has a cause! The phonetic nature of words is shaped by the experience of the people who produce them! There’s something delightfully kabbalistic about this.
The chili pepper quote promised a study on cultural learning of pain, so here it is:
Ken Craig has directly tested the relationship between cultural learning and pain. Ken’s team first exposed research participants to a series of electric shocks that gradually increased in intensity and thus painfulness. Some participants observed another person – a “tough model” – experience the same shocks right after them, and some did not. Both the participant and model had to rate how painful the shock was each time. The tough model, however, was secretly working for the experimenter and always rated the pain about 25% less painful than the participant did. Then, after this, the model left and the participants received a series of random electric shocks. For this new series of shocks, the participants who had seen the tough model rated them half as painful as those who didn’t see the tough model….
Those who saw the tough model showed (1) declining measurements of electrodermal skin potential, meaning that their bodies stopped reacting to the threat, (2) lower and more stable heart rates, and (3) lower stress ratings. Cultural learning from the tough model changed their physiological reactions to electric shocks.
I see a commenter on Quillette has already thought to connect this to telling people they should be harmed by triggers and microaggressions. But also note the connection to the the predictive processing model of perception.
Books like this are supposed to end with an Exhortation Relevant To Modern Society, so here’s Henrich’s:
Humans are bad at intentionally designing effective institutions and organizations, though I’m hoping that as we get deeper insights into human nature and cultural evolution this can improve. Until then, we should take a page from cultural evolution’s playbook and design “variation and selection systems” that will allow alternative institutions or organizational forms to compete. We can dump the losers, keep the winners, and hopefully gain some general insights during the process.
If that sounds familiar, it could be because it’s capitalism; if it sounds very familiar, it could be because it’s also the case for things like charter cities and seasteads; if it sounds super familiar, it could be because it’s also Archipelago.
And to finish:
Once we understand the importance of collective brains, we begin to see why modern societies differ in their innovativeness. It’s not the smartness of individuals or the formal incentives. It’s the willingness and ability of large numbers of individuals at the knowledge frontier to freely interact, exchange views, disagree, learn from each other, build collaborations, trust strangers, and be wrong.
Hopefully this means Henrich won’t be too angry that I just quoted like half of his copyrighted book without permission.