MindCORE Research Fellowship for Postdoctoral Scholars

/in

MindCORE seeks to recruit outstanding postdoctoral researchers for our Research Fellowship for Postdoctoral Scholars. Housed within the School of Arts and Sciences of the University of Pennsylvania, MindCORE is an interdisciplinary effort to understand human intelligence and behavior.

The MindCORE Fellowship is a springboard for early researchers who have recently obtained a PhD degree and aim to establish their own, independent research program that is consistent with MindCORE’s mission to understand the mind. Relevant disciplines include (but are not limited to) neuroscience, psychology, linguistics, philosophy, computer science, and cognitive science. Connections to other fields including math and physics are also encouraged. Fellows are expected to pursue collaborative relationships with faculty working across disciplines at Penn. These collaborations may take different forms from engaging in research projects with multiple faculty independently, to generating new collaborative research that brings multiple faculty together, to contributing to interdisciplinary discussion and reading groups. Find a list of current and former MindCORE Fellows here. MindCORE seeks to award 2 postdoctoral Fellowships per year. Positions may start as early as July 1, 2027.

Benefits

Fellows receive a competitive salary, relocation allowance, health insurance, plus a modest research budget of $20,000. Fellows also benefit from access to the greater community of academics including visiting scholars plus leading research facilities equipped with cutting-edge instrumentation all on an urban campus in a vibrant city. Because of the benefits of joining our Penn community, this is a Philadelphia-based (not remote) position. Fellows are invited to join regular working group meetings within their field plus career development workshops aimed at young researchers, and will be provided with a mentoring committee. Funding is provided in one-year terms renewable for up to three years.

Eligibility & Application

  • Applicants must have formally completed all requirements of the PhD degree and provide a copy of their diploma at the time of appointment (typically July 1, 2027-Jan 15, 2028). Penn has a five-year term limit for postdocs, so candidates ideally have 0-2 years of previous postdoc time.
  • Candidates must submit 1-2 page research statement that identifies at least three MindCORE faculty at Penn with whom the applicant would be interested in collaborating, along with a CV, and contact information for two referees (name and email addresses), and a cover letter addressing the applicant’s strengths with respect to the selection criteria below. Applicants should be mindful of the interdisciplinary backgrounds of the reviewers and make the proposals clear to readers.
  • Application documents should be saved with candidate’s Lastname_Firstname.pdf in one PDF file in this order: cover letter, research statement, CV, referees. (Note: We do not require recommendation letters at the time of application and only those candidates who are invited to interview will be asked to obtain letters.) Complete applications should be submitted through the form on our website. Questions not answered on our FAQ, can be emailed to pennmindcore@sas.upenn.edu
  • Applications due December 1, 2026.

Selection

All eligible and complete applications will be evaluated by the Selection Committee and are judged on the following criteria:

  • Scientific excellence
  • Potential to contribute to or lead within a diverse academic community
  • Scientific match and interdisciplinarity
  • Career potential

Sorcery: An Origins Story

/in

“The Sorcerer”, sketch based on Henri Breuil’s original drawing of art in the Trois-Frères cave, France.

– by Ronald J. Planer and Kim Sterelny

People in many traditional societies around the world respond to sickness, death, and a range of other serious misfortunes in a strange and strangely uniform way: they regard these events as ultimately caused by the occult agency of other humans. Moreover, this view is not just backward-looking; people actively fear occult attacks, and take countermeasures to avoid them. We call this psychological tendency “the agential view of misfortune”. Its wide occurrence is puzzling not just because the view is false, but also because the view is costly: socially, economically, and—one can only assume—psychologically. Michelle Stephenson (2022) provides a striking description of the view among the Mekeo of Papua New Guinea: “People seem almost obsessed by fears of harmful magic being used against them and no death or serious mishap occurs which is not attributed to the sorcerer’s powers. Villagers themselves say that they are preoccupied with such fears and that consequently they tend to be highly suspicious of each other’s motives, even of close relatives” (p. 129).

We take up this puzzle in our article, Rise of the dark shaman: A path from healers to sorcerers (2026). Our starting point is the inadequacy of existing explanations. The biggest problem, as we see it, is that none offers a plausible account of how the agential view of misfortune might establish in the first place. Once the view is established, its persistence is not mysterious. Acceptance by the parental generation strongly disposes acceptance by the next generation. The seriousness with which the parental generation takes the threat, and the costs they accept to reduce it testify to their real conviction. This is especially true to the extent that the local version of the view is enmeshed into the broader worldview of the culture—which it typically is. A person who questions the status quo may make themselves a target (indeed, they may be suspected of sorcery); at best, they will seem somewhat socially deviant. Moreover, accusations of sorcery and witchcraft are available as ready-made “social moves” for people to play against each other. Where and when people can deploy such accusations to improve or protect their social standing or that of their allies, we should expect they will do so.

We argue that the agential view of misfortune has its roots in shamanism (itself widespread). The prototypical shaman is someone who (i) is understood by others as possessing unique talents and skills for interacting with a supernatural world distinct from this one. (ii) They can obtain information no one else can, including knowledge of future events. (iii) Moreover, shaman are able to directly act in the spirit world on behalf of others; for example, they may locate and return the spirit of a sick person. But because they can act on behalf of others, they can act on their own behalf as well. They are players of in a dangerous game. (iv) To enter this world, shaman often subject themselves to excruciating physical rituals—often in the public eye—which may be supplemented by mind-altering drugs where available. They are the original “psychonauts”.

Shamanism primes the agential view of misfortune in two main ways. First, shamans constitute “living proof” of the possibility of sorcerers—of “dark shamans”. If they can wield their powers for good, it stands to reason they—or at least a counterpart—can wield them for evil. Many ethnographic accounts attest to group members’ impression that shaman are “morally ambiguous” in this sense. The second way is that shamans themselves stoke fear over sorcerous threats. As acknowledged occult specialists, shamans have an outsized role in shaping the occult lore and practices the next generation encounters, and by continuing to make the threat of sorcery salient to others, they ensure their ongoing place of social prominence in group life. Ethnography again shows that shamans (and similar occult specialists) enjoy prestige translating into benefits on the ground (e.g., being excused from normal social duties, having increased power in group decision-making). It is also common for shamans to receive direct compensation for services rendered (e.g., divination ceremonies). In some communities, these occult specialists also play forensic roles, revealing the source of occult attacks, and this too gives them social leverage.

We do not claim that shamans consciously “big up” themselves. Rather, we see the operation of a well-known and studied cognitive bias at work here—namely, the “myside bias”. Shamans have an implicit, unconscious tendency to bias the flow of information about the occult in ways that support and extend their importance. Our hypothesis is that this myside bias, operating cumulatively over cultural-historical time, leads to an elaboration of lore emphasizing the threat of harmful magic, and the shaman as defender and protector.

Our other main task in our article is to offer an account of how shamanism itself can establish. The core ideas we develop are these: (i) following an ingenious insight of 19th-century anthropologist Edward Burnett Tylor, we suggest that dreams and hallucinatory visions psychologically dispose people towards belief in immaterial spirits and a spirit realm. (ii) Once language evolved, these phenomenological raw materials were channeled into culturally shared views about spirits and the spirt world—in other words, a version of an animistic worldview was established. (iii) Proto-shamans were recognizable within this animistic milieu as individuals as who were especially willing to work with conscious-altering techniques that had already come to be associated with the spirit world, who were less risk-averse, and who were perhaps more prone to respond to these techniques.

The final step here is to explain how individuals claiming supernatural skills and talents for themselves gain credibility in the eyes of the broader group. We canvass a range of options in our article, but at an abstract level, we think cognitive biases are again crucial—especially our susceptibility to causal illusions. We are apt to turn noticed associations, and even the appearance of associations, into causal beliefs. When a highly salient event precedes a second, unusual or unexpected event, (an intense trance-inducing ceremony followed by the death of an enemy) humans are apt to conclude that the first caused the second. On our account, these biases play a dual role: they both work to convince proto-shamans of their powers while also convincing others of the efficacy of the displays and treatments these individuals put on.

These two parts of our argument—the development of the agential view of misfortune out of shamanism and the origins of shamanism itself—could be separated. In particular, one could couple the first to an alternative account of the latter. What we most want to insist on is that shamanism, however evolved, has a special role to play in midwifing the agential view of misfortune. Unlike belief in the omnipresent danger of others’ harmful magic, which is undeniably costly, belief in shamanism is plausibly beneficial to the broader group (if only for its placebo effect), or at the very least benign. Shamans did not displace more effective medical practitioners. This makes the task of explaining the origins of shamanism much easier.

Boiled down, what we propose is a (cultural) evolutionary model of belief in sorcery, witchcraft and allied phenomena in three main stages: first comes animism, then comes shamanism, then comes the agential view of misfortune.

For now, all we can claim to be in possession of is a “how plausibly” model. We take this to be progress, given the puzzling nature of the agential view of misfortune, together with the failure of previous explanations to deal with the unique challenge of the view’s origins. To convert our model into a “how actually” model will require testing, and we conclude our article with a number of predictions for future research. Stay tuned.

Ronald J. Planer and Kim Sterelny. Rise of the dark shaman: a path from healers to sorcery. Evolution and Human Behavior 47.5 (2026): 106912.

References

Stephen, M. (2022). Sorcery, magic and the Mekeo world view. In N.C. Habel (Ed.), Power, plumes and piglets, Australian Association for the Study of Religions Book Series, pp. 149-160.

Seeing Interest That Isn’t There: Four Routes to Sexual Overperception

/in

The same social cues can be interpreted as friendliness or sexual interest (Photo by James X on Unsplash)

– by Iliana Samara

Someone holds your gaze, laughs at your jokes, and suggests continuing the conversation over a drink. You might interpret their behavior as sexual interest, but they are only enjoying your company. Research on sexual overperception has often shown this pattern: men tend to infer more sexual interest from ambiguous behavior than women do (Abbey, 1982; Haselton & Buss, 2000). However, the psychological process underlying this pattern remains less well understood.

In a recent article in Evolution and Human Behavior (Samara, 2026), I address this issue and argue that the measures we commonly use to measure this pattern do not allow us to identify the underlying process. Work on sexual overperception measures how interested we think a person is in us (e.g., with rating scales) or how often one mistakes an interaction for flirtatious or romantic. However, patterns of higher interest ratings or higher frequency of mistaken interactions can arise through different psychological processes. Therefore, using the term “overperception” does not capture which process produced it.

Error management theory (EMT; Haselton & Buss, 2000) provides one influential evolutionary account for this pattern. EMT posits that when the consequences of two errors differ, selection may favour a decision policy that reduces the more costly mistake, even if this increases the other type of error. Applied to sexual interest, the notion is that ancestral men may have faced greater reproductive costs from missing a mating opportunity than from making an unsuccessful advance. Under these conditions, a tendency to infer interest from relatively weak evidence could have been advantageous for men.

We can identify four different routes that lead to this error. The first is cue discrimination: how well someone distinguishes friendliness accompanied by sexual interest from friendliness alone. If the cues are difficult to tell apart, errors can arise in both directions (inferring interest when it was not present or missing interest when it was present). The second concerns expectations: someone who starts with the expectation that sexual interest is likely may require less evidence before judging it as such. Such prior beliefs are dependent on the individual and might for example reflect previous experiences or beliefs about how likely sexual interest is in a given context. The third route concerns the perceived consequences of being wrong. Two people could have similar expectations and read the cues equally well but require different levels of evidence before indicating that another person is interested in them. For example, someone especially concerned about missing genuine interest may use a lower threshold than someone especially concerned about mistakenly inferring interest. Finally, the fourth route concerns the answer someone reports may differ from their private judgment, their reporting policy. A difference in reported judgments could therefore partly reflect differences in what people feel comfortable disclosing.

These processes can also occur together. Signal detection theory helps by separately estimating discrimination and the decision threshold (see Brandner et al., 2021, for an application to sexual overperception). However, even after estimating that threshold, we still need to explain how it arises, as a lower threshold could reflect stronger expectations of interest, greater concern about overlooking interest, or differences in reporting. Bayesian decision theory makes the contribution of expectations and error costs clear (Green & Swets, 1966).

To examine how well existing designs distinguish between these possibilities, I conducted a targeted audit of 54 empirical papers. Only five reported signal detection measures. None directly elicited expectations as estimates of how often interest was present, and none experimentally manipulated the payoffs attached to different errors. Many of these studies addressed other questions, including when overperception occurs and which factors predict it. Overall, the audit showed that their designs generally leave the underlying processes difficult to separate.

Future work could address this point by for instance, varying how often interest is present in a task, asking participants what prevalence they expect, and testing whether their beliefs change. Studies could also vary who will see an answer while keeping the available cues the same. Each manipulation would help test a different explanation, given it changes the beliefs or incentives it is intended to change.

The distinction between mechanisms can inform future interventions. If someone struggles to distinguish relevant cues, practice with feedback might help. If their expectations are poorly calibrated, feedback about how often interest is reciprocated may be more relevant. If perceived consequences or disclosure conditions drive the pattern, those would need separate attention. These are different intervention hypotheses, and the observed error pattern alone cannot show us which to pursue. For evolutionary research, identifying these processes would clarify how a proposed function is implemented and why behavior changes across situations.

Samara, I. Sexual overperception as a proximate identification problem. Evolution and Human Behavior 47.5 (2026): 106947.

References

Abbey, A. (1982). Sex differences in attributions for friendly behavior: Do males misperceive females’ friendliness? Journal of Personality and Social Psychology, 42(5), 830–838. https://doi.org/10.1037/0022-3514.42.5.830

Brandner, J. L., Pohlman, J., & Brase, G. L. (2021). On hits and being hit on: Error management theory, signal detection theory, and the male sexual overperception bias. Evolution and Human Behavior, 42(4), 331–342. https://doi.org/10.1016/j.evolhumbehav.2021.01.002

Green, D. M., & Swets, J. A. (1966). Signal detection theory and psychophysics (1st ed.). Wiley.

Haselton, M. G., & Buss, D. M. (2000). Error management theory: A new perspective on biases in cross-sex mind reading. Journal of Personality and Social Psychology, 78(1), 81–91. https://doi.org/10.1037/0022-3514.78.1.81

Possession is nine-tenths of the law – but maybe not for infants

/in

– by Erik K. Fonn

One important aspect of our social world is that we have intuitions that specific people have stronger claims to specific objects. We often refer to this as ownership. Getting this right is likely adaptively important: Taking the wrong things might be dangerous, and people might start to dislike you for it. But getting ownership right is in principle also difficult because these intuitions often bear on things that happened in the past: that is, how and by whom something was acquired. For instance, whether a person took an object by force or bought it likely matters for our intuitions about whether the object belongs to that person. But past events like these are often unavailable to us. So how are we able to form intuitions about who owns what?

Many studies suggest that humans, at least from early childhood onwards, rely on some mental shortcuts in order to make these ownership ascriptions. One such shortcut is prior possession. Simply put, this is the rule that whoever apparently had a resource first gets to keep it. And there are good evolutionary reasons that this is so. For instance, a person buying a house might start furnishing it according to their specific needs and preferences, making the house more valuable to themselves but not necessarily more valuable to everyone else. In this case, the house owner would tend to be more willing to defend their house than others are willing to fight to take it, and this general pattern has been documented across a range of animal species. Moreover, because losing possessions to others is costly, people might avoid those who do not respect prior possession and choose others as social and cooperative partners. Among humans, being a part of a social group and reaping the benefits of cooperation is indeed important to survive and thrive.

However, very little is known about how this prior possession rule actually emerges in human development: Is it something we learn from experience, or does it perhaps depend on cultural norms? Another possibility, which we sought to test, is that the prior possession rule is present in infants before they have had extensive experience with conflicts and before they have learned much about the world through socialization. In other words, the prior possession rule might serve as a backbone of human social understanding helping us to make sense of and navigate the social world from earliest childhood. Indeed, previous research suggests that infants have a very sophisticated understanding of the social world: Long before they can talk, infants distinguish agents from objects, predict who will win conflicts, and figure out who belongs to which group. Do infants, then, use prior possession to make sense of their social surroundings?

We approached this question in two ways: First, we asked whether infants avoid those who take, or attempt to take, things from others. Second, we asked whether infants expect those who hold prior possession to retain this possession when challenged. In order to address the first question, we showed 10–13-month-olds live puppet shows of one puppet possessing a toy box and another who did not. Subsequently, the one who did not have a toy box challenged the other. We varied across participants whether the one who held first possession lost the object, retained it, or whether the conflict was left unresolved. Finally, we tested whether infants preferred either of the two agents: They were presented with the two puppets on a tray and were allowed to choose between them. Across all the experimental conditions, infants were just as likely to choose the prior possessor as the contender. In other words, infants did not appear to avoid those who violate first possession.

Although infants did not systematically evaluate these social scenarios, they may in principle still have expectations that first possessors will win. Across two experiments we showed infants aged 10–18 months videos of a prior possessor (standing in the middle of an enclosure) being challenged by a contender. Infants saw both outcomes: either the prior possessor won the enclosure, or the contender did. If infants expect prior possessors to win, they should be surprised and look longer at the videos where the contender wins the territory. However, across both experiments, we found no evidence that they did. Yet, infants were able to understand the conflict: When we changed the size of the agents, so that one agent was much bigger than the other, infants expected the larger individual to win. In other words, infants were able to understand what was going on in the videos, but they did not care about who held prior possession.

These results did not align with our hypothesis that infants would use prior possession to make sense of their social world. Specifically, while prior research suggests that infants are able to track possession history in certain ways, for example distinguishing giving from taking, our results suggest that infants do not yet have the intuition that prior possessors have a stronger claim.

Of course, absence of evidence is not evidence of absence: We do not know if infants lack these capacities altogether, or whether we used the wrong methods to detect them. On the other hand, we found no effects of prior possession across two different measures, and infants did use physical size to predict who would win. This suggests that our methods were suitable to test infants’ conflict expectations, but that these expectations were not guided by prior possession alone. This asymmetry is puzzling: Predicting conflict outcomes from both size and prior possession would plausibly serve the same adaptive function: predicting who will win a conflict and so avoiding costly fights not worth having. Yet only physical size seems to guide infants’ conflict predictions. Whether and how infants come to use possession history to anticipate conflict outcomes still remains a question for future research.

Erik K. Fonn, Joakim H. Zahl, Bjørn D. Kristensen, Oliver Scott Curry, and Lotte Thomsen. Infants use physical size, but not prior possession, to predict conflict outcomes, and do not prefer prior possessors over contenders. Evolution and Human Behavior 47.5 (2026): 106952.

Why (And When) Are Young Children So Bad at Avoiding Germs?

/in

– by Serkan Özçakan, Florian van Leeuwen, and Joshua M. Tybur

Young children often behave as if they are indifferent to infection. They eat food from the floor, put dirty fingers in their mouths, and readily touch dirty surfaces. Importantly, young children do not just fail to take steps to avoid pathogens; they behave in ways that seem to approach pathogen risks. Paradoxically, all of this happens at a time when children are particularly vulnerable to pathogens, with their immune systems not yet mature. But why?

To answer this question, we first need to know when children begin avoiding pathogen threats and stop embracing them. The field currently lacks such a picture, for several reasons. Different relevant phenomena, including disgust, hygiene behaviors, and an understanding of contamination, have largely been studied separately. Measurement has sometimes been limited to questions asking parents to estimate the degree of disgust their child would experience during hypothetical events that the child may have never encountered. Child age is often binned into multi-year categories. Our study aimed to build a more comprehensive developmental map of pathogen avoidance by examining its multiple components simultaneously and at a high resolution of age.

To do so, we asked 3,050 parents in the UK and US whether they had observed each of 39 concrete behaviors in their child over the past month. Items concerned behaviors that increase pathogen exposure, such as eating food from the floor or touching dirty surfaces, as well as behaviors that reduce it, such as washing their hands or experiencing disgust. By recruiting parents of children from infancy to 10 years of age, we were able to estimate the prevalence of these behaviors across childhood in fine detail.

Our results showed that there was no single developmental timetable for pathogen avoidance. Instead, different behaviors emerged at different points in childhood. Disgust is a good example. Taste-related disgust and facial expressions of disgust toward food were already common by around six months. Vocalizations of “yuck” appeared around 20 months, followed soon after by more general disgust faces and reactions to sights, smells, and touch. The use of the word “disgust,” however, did not become common until around four years of age. In other words, disgust did not emerge all at once; its different components followed different developmental timelines.

We then asked whether behaviors with similar developmental patterns tended to group together. We observed four clusters of items with distinct developmental trajectories, which revealed two broader transitions around the second and fourth years of life. Around age two, a cluster of early exposure-increasing behaviors, including mouthing objects, became uncommon, while a cluster including assisted hygiene and early disgust responses became common. By age four, a second cluster of risk-prone behaviors, such as eating food from the floor, had also become uncommon, while more autonomous forms of pathogen avoidance, such as independent handwashing and toothbrushing, along with various expressions of disgust (e.g., saying “disgust” and reacting with disgust to things they saw or smelled), became common.

So, returning to the question we started with: why is pathogen avoidance so limited early in life despite high infection risk? Knowing when these behaviors develop gives us a starting point for thinking about why they follow this pattern. Many of the behaviors that are especially common early in life, such as mouthing inedible objects, licking dirty surfaces, and putting dirty hands in the mouth, allow children to explore their environment. This exploration may be important in at least two ways. It provides opportunities for learning about the world, but it also brings children into contact with microbes that may help shape the developing immune system.

The timing is particularly interesting here. The first 1,000 days of life are considered a key window for microbial colonization and immune development, when both the microbiota and the immune system are especially malleable (Zhu et al., 2025). Notably, this period broadly overlaps with the period in our data when behaviors that increase pathogen exposure were most common.

This overlap suggests that limited pathogen avoidance may reflect an adaptive trade-off between protection and exploration. High investments in pathogen avoidance from the beginning of life could limit opportunities for exploration at precisely the time when exploration may be especially important for learning and immune development. As children grow, this balance may shift toward protection. Behaviors that increase pathogen exposure become less common, while more organized forms of pathogen avoidance emerge. From this perspective, limited pathogen avoidance early in life may not simply reflect what young children cannot yet do. It may also reflect what they need to do: explore the world around them.

Özçakan, S., van Leeuwen, F., & Tybur, J. M. A developmental map of pathogen-avoidance behaviors. Evolution and Human Behavior 47.5 (2026): 106946.

References

Zhu, J., He, M., Li, S., Lei, Y., Xiang, X., Guo, Z., & Wang, Q. (2025). Shaping oral and intestinal microbiota and the immune system during the first 1,000 days of life. Frontiers in Pediatrics, 13, Article 1471743. https://doi.org/10.3389/fped.2025.1471743

How social connections shape our willingness to take risks

/in

– by Joon Hwang

Imagine you have two options: get $50 for sure or take a 50-50 chance of getting $100 or nothing. Both options have the same expected payoff ($50), but one might appeal to you more than the other. If you prefer to take that chance, you might be a risk-taker; if you prefer the guaranteed $50, you might be risk averse. You will notice that researchers use “risk” here to describe how much the possible payoffs vary, rather than harm or loss itself.

Now let’s add one more detail: you can count on family, friends, or neighbors to help you if you choose the risky option and get nothing. Would that make you more willing to take the risk? This is the question we examined in our recent paper in Evolution and Human Behavior. People have had to make decisions under risk throughout human history. Our ancestors could spend a day pursuing a large animal and return with plenty of meat or nothing at all. Today, someone might leave a steady job to start a business, hoping to earn more but knowing they could earn less or lose money.

Such unpredictability in everyday life is also closely linked to the evolution of human cooperation, particularly resource sharing. By sharing resources with others – giving when they have more and receiving when they have less – people can make their access to resources more reliable. The idea that sharing reduces variability of an outcome in this way has been widely studied under the “risk-reduction reciprocity” model of human cooperation. But in this study, we examined the decision itself. Having support to fall back on might do more than help you recover from a bad outcome. It might make you more willing to take a risk in the first place!

We worked with 140 adults from 82 households in Matlab, Bangladesh. In this rural area, farming and fishing, once the main sources of livelihood, have declined, while people’s dependence on markets and cash income has grown. The growing need for cash has made informal lending networks more important, while people continue to share food and household items with relatives and neighbors. We asked participants whom their households could borrow money (2,000 Bangladesh taka (BDT); about 21 USD), food, or household items from and whom they would lend to. From their answers, we mapped two social networks, with connections between households representing moneylending in one and the sharing of food and household items in the other. We also measured their willingness to take risks by asking them to choose between a guaranteed 50 BDT (about 0.52 USD) and a chance of getting 100 BDT (about 1.05 USD) or nothing, with the odds varying across rounds (see the figure below).

We found that different forms of social support were linked to greater or lower willingness to take risks, depending on the type of support and how people were connected. Participants were more willing to take risks when their households were part of tightly connected moneylending networks. This pattern is called “clustering”: the people you are connected to are also connected to one another. In moneylending networks, this could keep money circulating within a cluster. For example, money you lend to one household may later reach another household that can help you when you need it. Having financial support ties within this tightly connected cluster could help you cover a large expense or recover from a business loss. As a result, people may be more willing to take risks when they have reliable financial support to fall back on.

In contrast, participants who could turn to more households for food and household items were less willing to take risks. We had expected this support to encourage risk-taking too, but food and household items are worth less than cash loans and may not provide enough of a buffer against large losses. Instead, this support may act as informal insurance, helping people meet their everyday needs so they do not have to take risks out of desperation.

Although our results do not establish causal relationships, they help us consider how the social connections around us and the resources available through them might provide cues for responding flexibly to uncertainty. These cues may make a risky choice easier to afford (e.g., financial clustering) or reduce the need to take that risk at all (e.g., sharing food or household items). Seen this way, human systems for risk management may have evolved within networks of cooperation, where the structure of relationships shaped not only how resources were distributed to reduce variability in outcomes, but also how people made decisions under risk in the first place.

Should we worry, then, that helping people recover from losses might encourage them to take risks recklessly, a problem known as “moral hazard”? We suggest looking at it another way: innovation also involves risk, whether someone is trying a new farming method or starting a business. People may be more willing to try a new idea if others in their social networks can help cover the costs of failure, while the prospect of sharing in the gains through social networks may give those people a reason to offer that help. Thus, innovations we often credit to individual ingenuity may owe their success to the social environment that lets people take risks and recover from failed attempts. If so, we may also want to reconsider how the gains from successful innovations are shared, both to recognize that collective contribution and to sustain the cooperation that supports future innovation.

Joon Hwang, Nurul Alam, and Mary K. Shenk. Social capital is associated with both risk-taking and risk-avoidance in rural Bangladesh. Evolution and Human Behavior 47.5 (2026): 106923.

Averageness Isn’t Enough: What Actually Makes a Body Look Attractive?

/in

Photo by SHVETS Production / Pexels (faces blurred for anonymity).

– by Zihan Zhu, Lan Chen, and Ying Hu

Think about the last time you saw someone and thought, “They have a great figure.” What were you reacting to: something ordinary, or something that stood out? For decades, research on faces has pointed to a clear answer: averageness is attractive. Composite faces, blended from many individual faces, are consistently rated as more appealing than the originals, possibly because looking average signals health or biological quality. Body shape also varies along many dimensions, curvier or straighter, more or less muscular, and so on, yet it has received far less attention than faces. Does the same rule extend to bodies? Is a body more attractive simply because it sits closer to the average, or is something else going on?

To find out, we conducted two studies. In the first, 72 participants, 36 in the United States and 36 in China, rated 140 computer-generated 3D bodies (70 female, 70 male) that varied naturally along ten underlying dimensions of body shape. For each body, participants gave two separate ratings: how typical it looked (how representative it was of bodies in general) and how attractive it looked.

The two ratings were positively related. Bodies rated as more typical were also, on average, rated as more attractive, a pattern observed among both American and Chinese participants and for both female and male bodies. So, are typicality and attractiveness just two names for the same judgment? Not quite. When we examined what specifically predicted each rating, a different picture emerged.

We tested two competing explanations. One is that a body looks typical or attractive to the extent that it sits close to the population average — call it the “distance from average” rule. The other is that judgments depend not just on distance, but on which specific direction a body deviates in — call it the “specific direction” rule. Typicality was more closely tied to the first rule overall: bodies closer to the average tended to look more typical. Attractiveness, by contrast, was better captured by the second rule, with simple distance from average generally providing a weaker explanation. What mattered more was direction: bodies rated as highly attractive showed patterned departures from the average, including visible differences in curvature or muscularity, rather than simply being the most average bodies. This broad pattern was observed among both American and Chinese raters, even though the two cultures do not always agree on what an ideal body looks like.

To test this more directly, we ran a second, preregistered study using caricature and morphing. Using the ratings from Study 1, we built two composite bodies — one by averaging the five bodies rated most typical, and the other by averaging the five bodies rated most attractive. We then exaggerated the difference between them to create two caricatures: a “super-typical” body and a “super-attractive” body. Morphing gradually between these two caricatures produced an 11-step continuum, and a new sample of participants rated bodies at each step, some judging attractiveness, others typicality. We ran this twice, once letting body height vary naturally and once holding height constant, to make sure that height itself was not driving the results. Across a total of 353 participants, both versions, and both cultural samples, the same pattern held: as bodies moved toward the attractive end, attractiveness ratings increased overall, while typicality ratings did not increase in parallel, instead remaining broadly stable or decreasing toward the attractive end. Becoming more attractive did not mean becoming more typical, whether or not height was held constant.

Put together, the two studies suggest that typicality and attractiveness are related but not interchangeable judgments. Both may draw on the same underlying map of possible body shapes, but they read that map differently. Typicality asks a simple question: how close is this body to everyone else’s average? Attractiveness asks a more selective one: does this body show the “right” departures from the average, along dimensions like curvature and muscularity that plausibly carry information about health or reproductive fitness? Most bodies near the average do not show pronounced versions of these features, which is why typical and attractive bodies usually go together. But when a body’s shape leans strongly into these specific directions, the two judgments can pull apart. That is not the same as saying that unusual bodies are attractive; deviation from the average alone is not enough to explain attractiveness. It is the direction that counts, not distinctiveness for its own sake.

So the next time someone strikes you as having “a great figure,” your reaction may not be just about how ordinary or unusual they look overall. It may also be about whether their shape leans into a handful of specific features — a bit more curve here, a bit more muscle there — that read as attractive regardless of how average everything else is. Averageness still matters. It is an important cue to typicality, and typical bodies are, on the whole, seen as reasonably attractive too. It is just not the whole story.

Zhu, Z., Fu, X., Chen, L., Liu, C. H., Parde, C. J., & Hu, Y. Averageness is not enough: Different perceptual codes for body attractiveness versus typicality. Evolution and Human Behavior 47.5 (2026): 106953

Children Doubt Magic but Believe in Miracles

/in

– by Ayse Payir and Paul L. Harris

Young children are typically seen as naturally credulous beings, endowed with a rich imagination and incapable of distinguishing fantasy from reality. Given that they believe in entities like Santa, create imaginary friends, and fear monsters under the bed, this common stereotype is hard to dismiss.

In line with this stereotype, an early claim in developmental psychology was that young children are prone to magical thinking. Piaget argued that children’s magical thinking arises from the lack of any boundary between their mind and the outside world, so that the content of their mental world—dreams, emotions, imaginings—spill over into their ideas about physical reality. He proposed that children’s animism—their tendency to attribute mental states to inanimate beings—and artificialism—their tendency to think that natural phenomena are created to serve a purpose—are two manifestations of this spillover. So, according to a young child, clouds “…come from the smoke that people make in their chimneys…to make it rain for the gardens,” and clouds “… move because they want to make it rain somewhere else.” (Piaget, 1929). Gradually, such magical explanations are replaced with naturalistic explanations based on cause and effect: The clouds move “…because of the wind. The air moves and pushes the clouds along.”

Are children prone to magical causality?

Piaget’s proposal was initially challenged by Huang (1930) and Mead (1932), who showed that children provide very prosaic explanations even for events that violate their expectations. For example, in one study Huang covered a tube filled with water with a piece of paper over the mouth and quickly turned it upside down. When the water did not spill out (due to air pressure), children explained this seemingly strange event with natural rather than magical causation, arguing, for instance, that there must something sticky on the paper or at the rim of the tube which prevents the water to escape. Similarly, when Mead (1932) asked Manus children in Papua New Guinea to speculate about the cause of some unusual outcomes—such as a canoe escaping its moorings overnight—children typically invoked explanations that are quite mundane: Perhaps the owner did not tie up the canoe properly.

Decades later Hickling and Wellman (2001) reached similar conclusions when they analyzed children’s spontaneous causal explanations across a range of phenomena (e.g., objects, animals, humans) in everyday conversations. These explanations were not always factually correct, but they were consistent with the relevant domain and rarely involved magic. For example, children made a clear distinction between living versus nonliving beings and offered biological explanations for the former but not for the latter.

In a more recent study, we found further evidence for this naturalistic outlook (Payir et al., 2022). We presented children with short stories in which the main character experienced a bad outcome. For example, in one story a farmer ended up with a poor crop. When asked how the farmer could have avoided this bad outcome, children invoked realistic rather than supernatural or miraculous possibilities—for example “he could have watered the plants more”—and they did this regardless of their religious background, be it secular or religious.

In summary, recent research portrays children as rational, down-to-earth thinkers rather than muddled fantasists.

Are children natural-born believers?

If children are truly skeptical about magical causes, why do they believe in supernatural forces, such as divine agency? Maybe children are ‘born believers’ who are predisposed to invoke supernatural agency. Within the Cognitive Science of Religion framework researchers proposed children’s endorsement of God’s extraordinary powers (Barrett et al., 2001) and their tendency to entertain an afterlife (e.g., Bering, 2006) as evidence for this provocative claim. However, later studies showed that children attribute no extraordinary powers to God before the age of 5 to 7 (Heiphetz et al., 2016; Lane, 2021; Makris & Pnevmatikos, 2007) and initially endorse an irreligious, biological account of death (Harris & Giménez, 2005; Lane et al., 2016; Watson-Jones et al., 2017). By implication, children are not ‘born believers.’

A compelling case for young children’s belief in the divine was provided by Kelemen. Drawing on Piaget’s argument of young children’s artificialism (1929), Kelemen (2004) claimed that children have a natural inclination to ‘promiscuous teleology’—to think that everything in the natural world is created for a purpose by a divine being. Kelemen provided persuasive cross-cultural evidence that children do indeed think teleologically, but only children exposed to religion invoke a divine agent in doing so (Kelemen & DiYanni, 2005; Schachner et al., 2017). Essentially, children’s tendency to invoke a divine creator appears to be based on cultural learning rather than any innate tendency.

How do children learn to believe?

If children eventually come to believe that God exists via cultural learning, what exactly makes them endorse this claim, which is hard to verify? Anthropologist Henrich (2009) argued that unverifiable claims are not convincing unless accompanied by behaviors that enhance their credibility, for example, actions that pose obvious costs to the believer—such as ascetism, tithing or fasting. Such actions speak louder than words. Supporting Henrich’s proposal, various studies with adults indicate a strong relation between parental displays of religiosity during childhood and conviction in adulthood (e.g., Gervais & Najle, 2015; Willard & Cingl 2017).

Further support for the key role of cultural learning comes from developmental research on the role of testimony in learning about hard-to-observe phenomena. Studies indicate that across the world, children’s and adults’ belief in such phenomena—be they scientific, such as germs, or religious, such as God—vary based on the community consensus surrounding those phenomena (Ma et al., 2024). Accordingly, even in highly religious communities like Iran, there is greater confidence in the existence of scientific entities—whose existence is generally taken for granted—as compared to religious entities—for which the community consensus is less secure. Studies analyzing parent-child conversations show that this differential confidence is reflected in parental talk: Even religious parents talk about religious entities with less certainty as compared to scientific entities, and such linguistic cues moderate children’s level of confidence, providing direct evidence for the transmission of beliefs via testimony (McLoughlin et al., 2021; 2023).

In sum, scientific evidence casts doubt on the stereotype of the child as “magical thinker,” and suggests instead that children are naturalistic thinkers when left to their own devices. Nevertheless, they learn about supernatural possibilities, such as divine agency and the afterlife, from others in their family and immediate community. Exactly why such cultural learning is so potent remains an intriguing question that we hope to tackle in future research. (1083 words)

Ayse Payir and Paul L. Harris. Children doubt magic but believe in miracles. Evolution and Human Behavior 47.5 (2026): 106920.

References

Barrett, J. L., Richert, R. A. & Driesenga, A. (2001). God’s beliefs versus mother’s: The development of non-human agent concepts. Child Development, 72, 50-65. https://doi.org/10.1111/1467-8624.00265

Bering, J. M. (2006). The folk psychology of souls. Behavioral and Brain Sciences, 29, 453–498.

Gervais, W. M., Najle, M. B., & Caluori, N. (2021). The origins of religious disbelief: A dual inheritance approach. Social Psychological and Personality Science, 12(7), 1369–1379. https://doi.org/10.1177/1948550621994001

Harris, P. L. & Giménez, M. (2005). Children’s acceptance of conflicting testimony: The case of death. Journal of Cognition and Culture, 5, 143-164.  https://doi.org/10.1163/1568537054068606

Heiphetz, L., Lane, J. D.  Waytz, A. & Young, L. L. (2016).  How children and adults represent God’s mind. Cognitive Science, 40, 121-144. https://doi.org/10.1111/cogs.12232

Henrich, J. (2009). The evolution of costly displays, cooperation and religion: Credibility enhancing displays and their implications for cultural evolution. Evolution and human behavior, 30(4), 244-260. https://psycnet.apa.org/doi/10.1016/j.evolhumbehav.2009.03.005

Hickling, A. K., & Wellman, H. M. (2001). The emergence of children’s causal explanations and theories: Evidence from everyday conversation. Developmental Psychology, 37, 668–683. https://doi.org/10.1037/0012-1649.37.5.668

Huang, I. (1930). Children’s explanations of strange phenomena. Psychologische Forschung, 14, 63-183. https://doi.org/10.1007/BF00403871

Kelemen, D. (2004). Are children “Intuitive Theists”?: Reasoning about purpose and design in nature. Psychological Science, 15, 295-301. https://doi.org/10.1111/j.0956-7976.2004.00672.x

Kelemen, D & DiYanni, C. (2005). Intuitions about origins: Purpose and intelligent design in children’s reasoning about nature. Journal of Cognition and Development, 6, 3-31. https://doi.org/10.1207/s15327647jcd0601_2

Lane, J. D. (2021). Constructing ideas of the supernatural. Journal of Cognition and Development, 22(3), 343-355. https://doi.org/10.1080/15248372.2021.1906679.

Lane, J. D., Zhu, L., Evans, M. E. & Wellman, H. M. (2016). Developing concepts of the mind, body, and afterlife: exploring the roles of narrative context and culture. Journal of Cognition and Culture, 16, 50–82. https://doi.org/10.1163/15685373-12342168

Ma, S., Payir, A., McLoughlin, N., & Harris, P. L. (2024). Scientific and religious beliefs are primarily shaped by testimony. Trends in Cognitive Sciences, 28(9), 792-803. https://doi.org/10.1016/j.tics.2024.04.014

Makris, N. & Pnevmatikos, D.  (2007). Children’s understanding of human and super-natural mind. Cognitive Development, 22, 65–375. https://doi.org/10.1016/j.cogdev.2006.12.003

Mead, M. (1932). An investigation of the thought of primitive children, with special reference to animism. Journal of the Royal Anthropological Institute, 62, 173-190.

McLoughlin, N., Cui, Y. K., Davoodi, T., Payir, A., Clegg, J. M., Harris, P. L., & Corriveau, K. H. (2023). Expressions of uncertainty in invisible scientific and religious phenomena during naturalistic conversation. Cognition, 237, 105474. https://doi.org/10.1016/j.cognition.2023.105474

McLoughlin, N., Jacob, C., Samrow, P., & Corriveau, K. H. (2021). Beliefs about unobservable scientific and religious entities are transmitted via subtle linguistic cues in parental testimony. Journal of Cognition and Development, 22(3), 379-397. https://doi.org/10.1080/15248372.2020.1871351

Payir, A., Corriveau, K. H., & Harris, P. L. (2026). Peaceful coexistence or inevitable clash? Confronting the claims of science and religion about death. Child Development, 1-17.

https://doi.org/10.1093/chidev/aacag035

Payir, A., Heiphetz, L., Harris, P. L., & Corriveau, K. H. (2022). What could have been done? Counterfactual alternatives to negative outcomes generated by religious and secular children. Developmental Psychology, 58(2), 376-391. https://doi.org/10.1037/dev0001294

Piaget, J. (1929). The child’s conception of the world. Harcourt Brace

Schachner, A., Zhu, L., Li, J., & Kelemen, D. (2017). Is the bias for function-based explanations culturally universal? Children from China endorse teleological explanations of natural phenomena. Journal of Experimental Child Psychology, 157, 29-48. https://doi.org/10.1016/j.jecp.2016.12.006

Watson-Jones, R. E., Busch, J. T. A., Harris, P. L., & Legare, C. H. (2017). Does the body survive death? Cultural variation in beliefs about life everlasting. Cognitive Science, 41, 455-476. https://doi.org/10.1111/cogs.12430

Willard, A. K., & Cingl, L. (2017). Testing theories of secularization and religious belief in the Czech Republic and Slovakia. Evolution and Human Behavior, 38(5), 604–615. https://doi.org/10.1016/j.evolhumbehav.2017.01.002

Casual sex and the paths to parenthood

/in

– by By Dmitry Grigoryev, Albina Gallyamova, and Elizaveta Komyaginskaya

If someone is more open to casual sex, would you expect them to have more children? The prediction has intuitive appeal: more partners could mean more opportunities for reproduction. In our recent study in Evolution and Human Behavior, however, people who were more open to uncommitted sex than others in their country tended to report fewer children.  Understanding this result requires separating several ideas that are often bundled together: life history trade-offs, mating effort, demographic context, and the pathways through which sexual relationships ultimately become births.

Life history theory starts from a general problem: organisms have limited time and resources, so investment in growth, survival, mating, and parenting involves trade-offs. In parts of the psychological life history literature, these trade-offs have often been organized along a fast–slow continuum, with earlier reproduction, greater mating effort, and openness to casual sex treated as features of a relatively “fast” strategy. But this way of grouping the traits together has become increasingly debated. Recent theoretical and empirical work has questioned whether mating-versus-parenting and offspring quantity-versus-quality are really expressions of the same underlying continuum, and studies do not consistently show that unrestricted sociosexuality goes together with earlier or greater reproduction. Whether these characteristics actually fit together is therefore an empirical question.

Working with an international consortium, we investigated that question using a preregistered analysis plan and data from 80,047 heterosexual adults across 92 countries. We examined sociosexuality: individual differences in openness to sex without commitment. Three questions assessed comfort with casual sex, sexual fantasies about someone just met, and the number of partners with whom respondents had had sex without seeking a committed relationship. We analyzed attitudes and desire together, separately from reported behavior. These questions did not directly measure the time or energy invested in finding partners or raising children.

We also constructed a country-level demographic index combining total fertility, adolescent fertility, and mortality before age 40. Higher scores described relatively faster demographic regimes, with higher fertility, more adolescent births, and greater early mortality. This index characterized national demographic conditions; it did not assign a life history strategy to each respondent. Participants separately reported how many children they had.

Because respondents were nested within countries, we took a multilevel approach and separated two questions that are easy to confuse. One asks whether people who are more sexually unrestricted than others in their own country have more children. The other asks whether countries whose respondents report higher average sociosexuality also have respondents with more children. These comparisons capture different sources of variation—and, as it turned out, they gave different answers. Within countries, people who were more sociosexually unrestricted than the average person in their country reported fewer children. This negative within-country association appeared both for attitudes and desire and for reported behavior. It remained after accounting for age and national reproductive timing, as well as national wealth, education, gender equality, and historical kinship institutions, such as extended-family residence and marriage rules. The associations were modest and stronger among women than men. They also persisted in analyses restricted to parents or respondents aged 40 and older.

A comparison across countries told a different story: countries where respondents reported more uncommitted sexual behavior also tended to have respondents with more children. We did not find the same country-level pattern for attitudes and desires. This does not mean that more sexually unrestricted individuals in those countries had more children. Within the same country, the individual-level association still ran in the opposite direction: more unrestricted individuals tended to report fewer children. This contrast is exactly why separating levels of analysis is so important.

People living in faster demographic regimes also reported more children after country-level adjustment. This is partly unsurprising since fertility itself contributes to our demographic index, but the broader pattern is still informative. In our paper, we treat these regimes as reproductive contexts rather than as psychological strategies carried by every individual: local reproductive timing, kin networks, norms, and institutions may all help make earlier and higher fertility more common and easier to sustain. Faster regimes initially appeared associated with lower sociosexuality, but that association was no longer statistically clear after adjustment for national wealth, education, gender equality, and kinship institutions. Country context also mattered for one specific part of sociosexuality: behavior. In faster demographic regimes, the negative association between a person’s unrestricted sexual behavior and number of children became weaker, especially among men. In other words, unrestricted behavior did not become a fertility advantage overall; rather, it was less strongly associated with having fewer children. One possible interpretation is that where reproduction begins earlier and sexual behavior, partnership, and childbearing remain more tightly coupled, mating behavior has a more direct route to births. Attitudes and desire showed little evidence of this contextual variation, and the behavioral variation was less clear in the older and parent-only samples.

The picture changed most when we took relationship status into account. Accounting for whether respondents were single, married, or in other relationship situations reduced the negative association for attitudes and desire by almost 90%; it was no longer statistically clear. The behavioral association weakened much less and remained negative. Relationships therefore deserve close attention in explaining how sexual dispositions and childbearing become connected.

Why might relationships matter so much? There are at least two directions of influence. Openness to casual sex may be related to the formation and stability of partnerships through which parenthood often occurs. But entering a stable partnership or becoming a parent may also change sexual interests and opportunities. Reliable partners and relatives can help sustain the prolonged demands of raising human children, so stable pair bonds don’t have to mean having fewer children; in some circumstances, they may make several births easier to support. And all of this depends on the surrounding social environment: access to сontraception, expectations about marriage, and available family support may further alter how sexual relationships are realized into births. Our cross-sectional data cannot tell us which of these pathways produced the associations we observed. Beyond the limits of the cross-sectional design, several features of the data also place some limits on how broadly these results can be interpreted. The data was self-reported, and respondents were about 31 years old on average, so many had not completed childbearing. About two-thirds were women. The response scale combined five or more children into one category. Country sample sizes were unequal, and cultural differences in willingness to report sexual experiences could affect comparisons. We left offspring survival and later reproduction unmeasured, so number of children at the time of the survey is therefore only an incomplete measure of lifetime reproductive success.

Even with those limitations, our results are difficult to reconcile with one simple story: that greater openness to casual sex is a straightforward marker of a single “fast” strategy that should also lead to more children. The multilevel approach shows where the pieces come apart. Faster demographic regimes are associated with higher fertility, while within those same societies, more sexually unrestricted individuals tend to report fewer children. They leave open how the underlying dispositions evolved. A useful next step is to follow people—and ideally couples—through partnership formation, first births, and subsequent births to see when and how mating dispositions are actually converted into parenthood. That may be a more useful question than asking whether sociosexuality is simply “fast” or “slow.”

Grigoryev, D., Gallyamova, A., Komyaginskaya, E., et al. (2026). Sociosexual orientation, life history strategy, and reproductive success: A large-scale cross-cultural analysis across 92 contemporary societies. Evolution and Human Behavior, 47(5), 106938.

Who really believes in magic?

/in

– by Neil Levy, Rob Ross, and Ryan McKay

Anthropology must steer a narrow course between twin dangers. On the one hand, there is the perennial risk of exoticizing other peoples. On the other hand, there is a danger of a failure to recognize genuine difference, rendering somewhat alien worldviews mere variants of our own. Nowhere do these risks loom larger than when it comes to understanding the attitudes of other cultures to magic.

On the one hand, if we interpret magic as an instrumental technology, we are faced with the problem of understanding how rational people could genuinely believe in its efficacy, given the evidence against it. On the other hand, if we deny that they are genuine believers, we risk obscuring their genuine differences from us. The same problem arises closer to home, at least for those of us who accept a squarely secular worldview: how can our neighbors or our families genuinely believe in the supernatural agents of religions?

Non-instrumental accounts of magic (and religion) focus on its roles in fostering a sense of belonging, community, and in satisfying existential needs, without postulating a wholehearted belief in it as an instrumental technology. These accounts are today powerfully challenged by instrumentalist accounts, that maintain that people are, in the main, genuine believers, and suggest that (nevertheless) universal cognitive biases and heuristics can explain a diversity of attitudes to magic, including genuine belief in its efficacy.

We take our inspiration from recent work in political science and psychology (some of it by us) that provides evidence that belief reports are not always sincere. People sometimes insincerely report believing bizarre claims (like some conspiracy theories) or politically relevant claims for a variety of reasons.

One motivation for insincere belief reports is to express an attitude. Political scientists call this expressive responding or partisan cheerleading. For example, people may report believing that the crowd at the first Trump inauguration was bigger than at Obama’s to express support for Trump, just as someone might report believing the Browns are the greatest team to express support for their side, rather than because one genuinely believes.

People may also report believing a conspiracy theory or engaging in a strange behavior (like drinking bleach to prevent COVID-19) for a different reason: to ‘troll’ experimenters or because they think it’s funny.

People may also sincerely but wrongly report their own beliefs. Asked to report what they believe about some question they’ve never explicitly considered before (“will Donald Trump invade Canada?”), people may sample a range of considerations that come to mind and construct a narrative. This report may not reflect their true attitude. The tendency to misreport an attitude may be particularly strong when the attitude is one that is fun to maintain and receives community support.

Most of the evidence, at least from political science and psychology, concerns the belief reports of people in WEIRD societies. We see no reason to think that the reports of people in non-WEIRD society are more likely to be sincere, and (when sincere) accurate. While the accusation that Margaret Mead’s informants were hoaxing them is sharply contested, there’s evidence that informants may troll anthropologists. Salamone quotes one Tiv informant’s ridicule: “When I read what the white man has written of our customs, I laugh, for it is the custom of our people to lie as a matter of course to outsiders, especially the white man.”

The evidence concerning expressive responding, trolling, and the construction of attitudes among WEIRD participants does not suggest that everyone is insincere or mistakenly reporting a belief. Rather, it supports what we call a motley of attitudes: some people are genuine believers, some are trolls, some are sincere but mistaken. The proportion of genuine believers varies from topic to topic, and a single individual may take a variety of attitudes across topics. We think that exactly the same is true of people in non-WEIRD societies when it comes to belief in magic. Some may understand magic symbolically, some may be sincere believers (or disbelievers) in it, and some may have no stable attitude at all.

Instrumentalist accounts offer compelling evidence that some people are really true believers. However, we point to evidence in the anthropological literature to argue that other attitudes exist alongside belief. We suggest that postulating a heterogeneity of attitudes is needed to make sense of many ethnographic reports without attributing a lack of rationality to informants. In his study of Burmese animism, Spiro notes that the beliefs of his informants are “marked by inconsistency, contradiction, and by what might be called cognitive looseness,” with different informants manifesting different degrees of belief ranging from wholehearted acceptance to outright skepticism. We suspect this sort of heterogeneity is widespread across many societies, including our own.

When Evans-Pritchard attributes inconsistent beliefs among the Azande and a blatant failure to “perceive the contradiction” between their beliefs, he risks not only making them alien, but failing to recognize their rationality. While we acknowledge the risk we run of the reverse mistake — of assuming that people from “exotic” societies are entirely like ourselves — we think our suggestion that people of all societies take a motley of attitudes is both more respectful of our shared humanity and more faithful to the evidence.

Neil Levy, Robert M. Ross, and Ryan McKay. A motley of attitudes: belief, insincerity, and non-doxastic commitment in magical practice. Evolution and Human Behavior 47.5 (2026): 106945