Simon Columbus

@simoncolumbus.bsky.social

1008 Followers 446 Following

Lecturer, University of St Andrews | I work on cooperation, norms, institutions, & personality | http://simoncolumbus.com

Abstract: In social interactions, individuals frequently face trade-offs between personal and collective interests, ranging from zero-sum situations — where personal gain may come at others’ expense — to positive-sum situations, where joint efforts lead to mutual benefit. Punishment can sustain cooperation by deterring free riding (prosocial punishment), yet it can also provoke retaliation; when directed at cooperators (antisocial punishment), punishment may even undermine collective welfare. In this talk, I will present two cross-cultural studies examining the role of the cultural logic of honor in shaping how individuals navigate these tensions and adopt particular punishment strategies in situations involving conflicts of interest. Drawing on a 13-society online behavioral experiment (n = 3,371), we found that perceived societal norms of honor were associated with greater competition in zero-sum situations, but also greater coordinative efforts in positive-sum situations at both individual and societal levels. In a multi-lab experiment conducted in Egypt (n = 266), South Korea (n = 212), and the UK (n = 192), we found that honor values and norms were not directly associated with engagement in prosocial or antisocial punishment. However, when individuals experience unfair treatment, honor values and norms intensified antisocial retaliatory responses. Together, these studies show how honor, as a relatively understudied cultural construct, helps explain variation in cooperation and punishment across societies.
Abstract: Prosocial behavior is common and often socially rewarded (e.g., via liking, status, and trust). Yet prior research has found that if actors themselves also benefit from their prosocial behavior, then they are morally derogated: They are evaluated as worse than purely selfish actors. This tainted-altruism effect has been explained by the use of different counterfactuals for the evaluation of prosocial and selfish actors. Here I propose social rewards protection theory, which explains why evaluators use these different counterfactuals in the first place: Social rewards are treated as being reserved for costly prosocial actions. Claiming such rewards without incurring costs seems like cheating and thus deserves moral derogation. Accordingly, being transparent about the action’s costs and benefits prevents such derogation. I conducted six experiments (five preregistered) with Amazon Mechanical Turk (MTurk) workers in the United States and lab participants in Spain (total N = 4,732 adults). The findings provide support for the proposed functional explanation of tainted altruism, which also sheds light on related phenomena, such as overhead aversion and hypocrisy.
Abstract: Accurately measuring social norms is crucial both for explaining cooperation and prosociality and for designing effective policy measures that support these. While established laboratory-based approaches represent “gold-standard” approaches, their practicality and widespread use are limited by the restrictions involved in their implementation. In contrast, non-incentivised survey measures are simple and widely used yet remain unvalidated, leaving it unclear what they truly capture. We address this gap by validating a popular set of non-incentivised measures against incentivised benchmarks across two diverse contexts—one WEIRD (Italy) and one non-WEIRD (India)—using student and general-population samples (N≈1,400) in laboratory and lab-in-the-field environments. Every measure is validated in seven settings representing a range of interaction domains including cooperation (e.g. Public Goods Game, Dictator Game). From these results, we assemble a menu of validated survey measures, highlighting their respective strengths and offering guidance for best practice across different research contexts and constraints.
Abstract: Repeated interaction is a key mechanism driving the evolution of cooperation, generosity, extortion, and other social behaviors. These behaviors are typically studied using two-player repeated games, where the two players act simultaneously in every round. In this paper, we begin with a standard two-action repeated game, but transform it into a leader–follower game. We do this by forcing one player to indicate their impending move, or to openly commit first in each round. We show that this simple transformation provides new and useful analytical insights. In particular, we use this transformation to derive expressions for the main classes of simultaneous-move Nash equilibria of memory-1, when payoffs are discounted. We show that these equilibria also remain stable under the leader–follower structure. We prove that this reassuring property does not always extend to games with more than two actions or to equilibrium strategies with longer memory. Overall, we establish a novel connection between repeated simultaneous-move games and repeated leader–follower games in evolutionary game theory. This approach offers fruitful means of investigating how prediction and anticipation play a role in stable social behavior.
Abstract: Social norms are unwritten rules that guide human behavior and social interaction. They play a central role in promoting cooperation and maintaining cultural diversity. Yet because norms pressure people to do what others do, they may also sustain practices that are inefficient or harmful. In this talk, I present insights from a mathematical model and a field study that together examine whether and how such norms can persist. The model challenges the idea that arbitrary or harmful norms can be maintained solely through social pressure. It shows that norms with continuously varying behavioral options cannot be maintained at ecologically or psychologically suboptimal equilibria, suggesting that norms adapt more readily than previously assumed. I then present evidence from a field study among the Derung, small-scale horticulturalists in Yunnan, China. Derung farmers continued to follow a traditional cooperative farming harvest division norm after market integration. They did so even though they recognized its lower ecological payoff and expected no social disapproval for deviating from it. Rather than payoff calculations, Derung farmers adhered to the norm out of a desire to follow tradition and common practice. However, five years later the norm had changed, consistent with the model’s prediction that norms with continuously varying options eventually adapt to ecological conditions.
Abstract: If only those behaviours evolve that increase the actor’s own survival and reproductive success, then it might come as a surprise that cooperative behaviours, i.e., providing benefits to others, are a widespread phenomenon. One possibility to explain these behaviours is reciprocal cooperation, i.e., helping those that were helpful before. Still, the evolutionary origin of it is largely unclear, mainly because it is believed that - in contrast to humans - other animals do not exchange help reciprocally. Yet, I propose that reciprocity is likely to be widespread. In my talk, I will first shed light on the mechanisms of reciprocal cooperation in Norway rats (Rattus norvegicus). In a series of studies, my colleagues and I have demonstrated that Norway rats reciprocally exchange goods and services between and within different commodities and that reciprocal cooperation is not necessarily cognitively demanding. Furthermore, to understand the evolutionary origins of human reciprocity, and whether it is shared with other animals, I will then discuss evidence for reciprocity in non-human primates, which are our closest living relatives. A thorough analysis of the findings showed that reciprocity is present and not confined to unrelated individuals. Based on my findings, I conclude that reciprocal cooperation in non-human animals is present but largely neglected and not restricted to humans.
A screenshot of a Zeit article saying Irankrieg
Abstract: Human societies are characterized by large-scale cooperative behavior, even among unrelated strangers. Why? One plausible answer is indirect reciprocity: If I help a stranger, I might gain a good reputation, increasing the likelihood that others will help me in turn. But under what circumstances can indirect reciprocity facilitate a stable, cooperative status quo? Can these circumstances—for example, a high level of agreement about reputations—evolve in the first place? For that matter, what rules for assigning reputations can evolve? I use adaptive dynamics to attain simple, intuitive answers to these questions. Specifically, I study the evolution of cooperative strategies, gossiping behavior, and even moral assessment rules themselves by parametrizing them as continuous traits. I confirm some classical results in the theory of indirect reciprocity but challenge others. Cooperation is easily sustainable, provided reputational views are sufficiently correlated. But when reputations are a matter of private judgment, there are no stable cooperative equilibria at all. Gossip can facilitate a high level of agreement, but echo chambers are difficult to overcome. Crucially, while many assessment rules can facilitate cooperation, almost none of them can actually evolve. The main exception is Stern Judging, which mandates that bad individuals be punished––a rule that is, sadly, highly intolerant of disagreement. I conclude by discussing implications for polarization, moral philosophy, and the discourse. Adaptive dynamics provides a mathematical framework not only for establishing a baseline theory of indirect reciprocity but also for incorporating the foibles and peculiarities of real human societies.
Abstract: What motivates people to contribute to Wikipedia? Have you heard of Nupedia? Why do software engineers contribute code to open source projects? Why do technology companies release their large language models as "open weight"? Why has participation declined on Stack Overflow since the release of ChatGPT, but not on the programming subreddit? Why do some crowdfunding sites succeed while others fail? Underlying each of these questions is a more fundamental one: how can social scientists, computer and information scientists and practitioners use social science theories combined with information technology to increase contributions to public goods? To explore this question, my book, "Digital Public Goods", introduces (1) case studies from the technology sector to set up the problems; (2) social science theories of motivation, production, informational nudges, identity and teams, goal setting, and reputation, together with their field implementations using information technologies, as solutions; and (3) experimental evaluations of these theories and solutions. The goal of this book is to merge the latest research in the scholarly literature with successful practices used by digital platforms to provide scholars, students and practitioners with the intellectual tools they need to develop an understanding of digital public goods provision.
Abtract: Prosocial behavior is key to addressing global collective-action challenges. Two prominent tools to promote prosociality are incentives and the use of conditional preferences, yet existing evidence is mixed and largely based on WEIRD populations. We study how incentives and conditionality shape prosocial behavior across diverse cultural contexts, and how cultural characteristics moderate these effects. 15,600 participants from 42 countries completed an online experiment in which prosociality was measured as the donation of earned bonus points to international charities (e.g., UNICEF). In the incentive condition, donors received additional points worth half of the donated amount. In the conditionality treatment, participants could condition their donation on the share of other donors in their country. Preregistered analyses show that incentives increase prosocial behavior overall, but that their effectiveness varies across societies. Incentive effects are stronger in societies where the reputational costs of benefiting personally from prosocial acts are lower, that is, where incentivized donors are perceived as relatively more likable. Conditional giving also increases prosocial behavior, yet the effects are larger in societies characterized by a higher prevalence of conditional preferences and stronger generalized fairness concerns. Together, these findings highlight that the effectiveness of behavioral interventions depends systematically on cultural context, underscoring the importance of moving beyond WEIRD samples not only for advancing theory, but also for culturally tailored policy design.
List of speakers: 2026-01-30 | Caroline Graf (University of Zürich) 2026-02-13 | Yan Chen (University of Michigan) 2026-02-27 | Taylor Kessinger (University of Pennsylvania) 2026-03-13 | Manon Schweinfurth (University of St Andrews) 2026-03-27 | Minhua Yan (Institute for Advanced Study Toulouse) 2026-04-10 | Saptarshi Pal (Harvard (University) 2026-04-24 | Aron Szekely (UCL) 2026-05-08 | Sebastian Hafenbrädl (IESE Business School) 2026-05-22 | Shuxian Jin (Peking University & University of Sussex) 2026-06-19 | Sara Constantino (Stanford University)
Abstract: A necessary prerequisite for the accumulation of beneficial knowledge, or ‘cumulative cultural evolution’, is the sharing of information via social learning. Yet little work in the field of cultural evolution has examined the mechanisms that support information sharing in the face of exploitative information free-riding and information hoarding. We ran a series of online interactive experiments (N = 716) combined with computational reinforcement and social learning models to test whether the mechanism of reputation-based partner choice can effectively support information sharing. Participants in groups chose whether to (i) engage in costly innovation and (ii) whether to share the resultant knowledge. Sharers received increased reputations for sharing and participants could use reputations to select recipients of knowledge. Study 1 found, in participants from the UK and China, high levels of information sharing but only weak partner choice. Study 2 replicated this finding with various methodological improvements. Study 3 showed that participants used partner choice to exclude non-sharing artificial bots from receiving information. Study 4 found that when partner choice was impossible, preferences for information sharing declined to levels comparable to non-sharing. Overall, our findings provide tentative positive experimental evidence that partner choice can facilitate information sharing and enable cumulative cultural evolution.
Abstract: The management of shared natural resources often hinges on local cooperation, yet the social networks linking communities across space can both complicate and enable collective action. This talk examines how long-distance friendships—relationships that bridge community boundaries—shape trust, cooperation, and participation in fisheries management in coastal Tanzania. Drawing on theoretical models and multi-site ethnographic and survey data from the Tanga Sociality and Fisheries Project, I show that long-distance friendships provide key forms of material support similar to local friendships, suggesting they are a resilient social strategy for managing ecological and economic risk while fostering interdependence across villages. In turn, people with more long-distance friends express greater trust in non-local partners and are more likely to engage in collective monitoring and reporting of destructive fishing practices. Finally, I introduce a new measure, “long-distance orientation,” capturing individual variation in the motivation and capacity to form and sustain these ties, highlighting psychological and structural factors that facilitate cooperation at a distance. Together, these findings suggest that long-distance relationships can serve as a foundation for scaling up collective action and designing management systems that bridge community boundaries.
Abstract: Gossip is an ever-present feature of our social interactions. We gossip to update one another about the others in our social environment. Our persistent gossip shapes perceptions and behaviors that are key to forming, keeping, and severing social relationships. These uses of gossip have given rise to the perspective that gossip has cooperative functions. This entails that people gossip to provide information about how to (not) behave and who can(not) be trusted to cooperate. This informs people about others’ reputations, which can serve to make better decisions about whom to cooperate with and whom to avoid. First, I will present research that supports this perspective based on experience sampling, scenario studies, and behavioral experiments. I find that people use reputational information provided by gossip to shape how they think about the targets of gossip. These gossip-based impressions guide receivers’ cooperative behavior toward targets: they cooperate with people they learn are likely to be cooperative and avoid people who they learn are likely to violate norms. Second, I focus on the puzzle of gossip as key for a cooperative social life while also being commonly perceived very negatively. Gossip is needed but could lead to costs for gossipers that motivate refraining from gossip. Based on ongoing research using behavioral experiments, I will present the first steps towards understanding the puzzle of the costs and benefits of gossip. Receivers of gossip support gossip systems by punishing harm to the reputation-system from not contributing gossip or spreading false information and by rewarding accurate gossip.
Abstract: Successful cooperation depends on mentalization—the ability to infer and adapt to others’ beliefs, intentions, and strategies. Yet, most existing studies and models treat mentalizing as a static trait rather than an adaptive, dynamic process unfolding within interactive contexts. In this talk, I will present a modelling framework and experimental paradigm that capture adaptive mentalization: the capacity to continuously update one’s internal model of another person’s behavioral sophistication and to adjust one’s own strategy accordingly. Using a social strategic game combined with computational modeling and fMRI, we develop and validate a Bayesian model indexing how effectively individuals can adapt their mental models in response to changing partner behavior, and we identify a multivariate neural signature that can predict this adaptiveness of mentalization. This neural pattern generalizes across individuals and closely tracks the flexibility of mental-state inference during interaction. Applying this paradigm to individuals with varying levels of autistic traits, we find that reduced adaptive mentalization —both behaviorally and neurally—correlates with higher trait expression. These results reveal how subtle alterations in the adaptiveness of mentalization can influence the dynamics of social interactions. By grounding social interaction in formal models of belief-updating and strategy adjustment, our work advances a more mechanistic understanding of how humans coordinate minds in social contexts—and how individual differences in these computations shape social behavior.
A scene from Raphael's The Triumph of Galatea: A cupid, labelled 'Psychologist', shooting an arrow labelled 'Causality'. The cupid's privates are censored because Bsky is run by prudish robots.
A scene from Raphael's The Triumph of Galatea: A cupid, labelled 'Psychologist', shooting an arrow labelled 'Causality'.
Abstract: Authorities, from parents of toddlers to leaders of formal institutions, use punishment to communicate disapproval and enforce social norms. However, punishment is often interpreted in contrasting ways: at times it teaches and enforces norms, but at other times it fails or even backfires, undermining the authority’s legitimacy. We show that these seemingly contradictory effects of punishment can only be understood by considering the cognitive processes in the minds of human observers of punishment. In a series of experiments, we showed the systematic variability in how adults and children interpret punishment. We developed a formal cognitive model, derived from a standard model of how people make sense of each others’ actions (Inverse Planning), that captures and explains these inferences quantitatively and parsimoniously. This work reveals the rational logic behind how people learn from punishment, and a key constraint on the function of punishment in establishing shared social norms.
Abstract: Mitigating the increasing threat of climate change poses a uniquely challenging cooperative problem. Multiple strategic features of climate change impede coordinating around successful action. For example, there is uncertainty about the timing and dynamics of climate change impacts. Different actors vary in their historic contribution to the problem and vulnerability now. There are many solutions available to mitigate climate change, but they differ in their effectiveness and probability of success. And, compounding all of this, there are asymmetries in information between elite decision-makers and the publics that they govern. In this talk, I argue economic games are well-suited to understand both how these strategic features delay successful climate action and provide insights into how to overcome these problems through public policy. Focusing on the case of governing emerging mitigation technologies, I will present the empirical evidence that people are willing to pay costs to avert climate disaster. While mistrust between the public and policymakers erodes cooperation, I will also present solutions that emerge from these experiments. I conclude by discussing how insights from the lab can inform real-world messaging and policymaking to encourage successful climate change mitigation.
A view from a Swiss alp near the Aletsch Glacier towards the Matterhorn.
A photo of a table. On it lays Banu Mushtaq's collection of short stories, Heart Lamp, next to an Aperol spritz.
Abstract: This paper develops a theoretical framework to examine how normative expectations--what one considers others should do based on their fairness perceptions--in shaping the behavior of players with reciprocal preferences in a public goods game. Standard reciprocity models typically assume uniform, exogenously determined, and moderate expectations. By contrast, I allow players to (i) vary in how demanding they are of others, (ii) disagree about what counts as fair, and (iii) form self-centered normative expectations endogenously. In noncooperative play, strong reciprocal concerns transform the payoff structure from a material payoff dilemma into a coordination game in utilities, with one exception. In coalition formation, three central results emerge: (i) the empty coalition is always stable, (ii) the grand coalition becomes stable once reciprocal concerns are sufficiently strong, and (iii) partial coalitions can be stable under specific conditions. Exogenous expectations create distinct effects on cooperation: lower expectations facilitate the stability of the grand coalition, whereas higher expectations increase the size of partial coalitions, potentially leading to a majority coalition. This latter finding is more consistent with empirical observation.
The opening lines to Garth Greenwell's Small Rain: "They asked me to describe the pain but the pain defied description, on a scale of one to ten it demanded a different scale."
A photo of a fish on a balcony. *My* balcony.
Speaker list for the Autumn 2025 Cooperation Colloquia: 2025-09-12 | Doruk İriş (Sogang University) 2025-09-26 | Talbot M. Andrews (Cornell University) 2025-10-10 | Setayesh Radkani (MIT) 2025-10-24 | Christian Ruff (University of Zürich) 2025-11-07 | Bianca Beersma (Vrije Universiteit Amsterdam) 2025-11-21 | Kristopher M. Smith (Washington State University) 2025-12-05 | Alex Mesoudi (University of Exeter)
Two people standing between the pillars of Buddo Rock, a sandstone rock on the coast of Fife near St Andrews.
Abstract: Indirect reciprocity is a key mechanism to understand the evolution of cooperation among unrelated individuals. By helping someone, individuals may improve their reputation, which may be shared in a population and change the predisposition of others to reciprocate in the future. The reputation of individuals depends, in turn, on social norms that define a good or bad action. Over the years, theoretical and empirical research has unveiled many features of cooperation under IR, exploring norms with varying degrees of complexity and information requirements. Often, however, it is assumed that 1) reputations spread for free, 2) individuals can adopt arbitrarily complex norms and 3) populations are fully composed by homogeneous individuals. In this presentation, I will discuss a way to quantify the complexity of norms and strategies in indirect reciprocity and the challenges introduced by complex rules and costly reputation sharing. We will observe that cooperation can prevail with simple norms and costly reputation sharing, provided that individuals can anticipate the reporting intentions of their opponents. I will conclude by discussing the potential of indirect reciprocity to sustain cooperation in populations that include both humans and artificial agents.
Abstract: Influential theories have proposed that impersonal prosociality varies across human societies. At the same time, different societies may rely on distinct mechanisms—guilt and internalized norms versus shame and external pressures—to promote prosocial behavior. We examine these mechanisms in a pre-registered experiment with 7,978 participants across 20 culturally diverse countries. Participants made monetary allocation decisions between themselves and strangers. To examine how guilt and shame affected prosociality, we introduced two variations to this standard decision-making task. First, to activate guilt, we varied information about the consequences of participants’ decisions (full versus hidden). Second, to activate shame, we varied the observability of participants’ decisions (public versus private). Additionally, we measured guilt- and shame-proneness at the individual and country levels. We find robust evidence for guilt-driven prosociality and willful ignorance across all countries. Allowing individuals to avoid information about the consequences of their actions substantially decreases prosociality. Guilt-prone, rather than shame-prone, individuals are more responsive to the presence of information about the negative consequences of their actions. In contrast, we find that making decisions observable among strangers has negligible effects on prosociality. Our study provides a first comprehensive investigation of how guilt and shame affect prosociality around the world.
Abstract: In Weisel and Zultan (2016, 2021a, 2021b, 2022), we claim that whether intergroup conflict is perceived as a threat to the individual or to the group informs the behavioral responses to conflict. In this talk, I will provide an overview of this research program, and share some background and ruminations beyond the information contained in the published papers.
A view over the River Tay from Dundee.
Abstract: Evolutionary theories have revealed various mechanisms that promote cooperation in biological organisms, and humans are no exceptions. The modern theory classifies those mechanisms into two major categories: one is kin selection, and the other is reciprocity. Kin selection means that the benefit of cooperation is passed to a relative who shares the same gene and therefore a gene that encodes cooperation is favored by natural selection. Reciprocity means that the benefit of cooperation is later given back to the cooperator in a reciprocal manner. In this talk, I will talk about those two mechanisms and how they can explain the evolutionary origin of human cooperation. First, I discuss a game-theoretic model that hypothesizes that modern human cooperation has its origin in our ancestral time when kin selection strongly operated. Second, I will talk about indirect reciprocity, which is reputation-based cooperation, and explain my recent theoretical finding that reciprocal cooperation can be robustly maintained even if reputation information is not public.
Abstract: The open science movement has gained significant momentum over the past decade, with pre-registration and the use of pre-analysis plans being central to ongoing debates. Combining observational evidence on trends in adoption with survey data from 519 re-searchers, this study examines the adoption of pre-registration (potentially but not necessarily including pre-analysis plans) in experimental economics. Pooling statistics from 19 leading journals published between 2017 and 2023, we observe that the number of papers containing a pre-registration grew from seven per year to 190 per year. Our findings indicate that pre-registration has now become mainstream in experimental economics, with two-thirds of respondents expressing favorable views and 86% having pre-registered at least one study. However, opinions are divided on the scope and comprehensiveness of pre-registration, highlighting the need for clearer guidelines. Researchers assign a credibility premium to pre-registered tests, although the exact channels remain to be understood. Our results suggest growing support for open science practices among experimental economists, with demand for professional associations to guide researchers and reviewers on best practices for pre-registration and other open science initiatives.
Abstract: When making decisions, individuals do not only take into account their own payoffs but often additionally consider the consequences for others. Models of social preferences (e.g., social value orientation) aim to describe exactly how outcome for the self and for others are integrated. Models widely differ concerning whether and how various factors are taken into account. They contain for example different relative weights for the own and the other outcome, take into account inequality (i.e. differences in outcomes for self and other) and/or efficiency concerns (i.e. the sum of outcomes). Some models, such as Jenkins et al. (2018), also differentiate between cases of advantageous and disadvantageous inequality. Jenkins et al. suggested a Social Perception-Weighted (SPW) model of social valuation, which takes into account perceived competence and warmth from the stereotype content model and show its superiority in predicting behavior. We conduct a comprehensive model comparison by (i) applying SPW to cross-national interactions, (ii) testing SPW against competing models, and (iii) extending SPW by including further variables. We analyze choices from an incentivized, multi-national study (N = 6,182). In the experiment, individuals from 25 nations allocated money between themselves and individuals from the own and each other nation. Stereotype perception, perceived conflict between nations and further factors were measured. Based on these model comparisons, we identify relevant structures and factors and suggest an extended model of social preference for cross-national interactions taking into account social perception.
Abstract: Whether people actively intervene in socially harmful behavior, which factors predict this decision, and whether the decision can be influenced has been little studied so far. Here, we investigate these questions using the example of cigarette littering, a behavior that is ecologically, economically, and socially harmful, yet widespread. The results, which are based on a survey experiment combined with field experimental investigations from Germany, show that the willingness to intervene is generally rather low, especially when the litterer is a stranger, people lack intrinsic motivation, are accustomed to cigarette littering, and do not have small children. Expectations about how approached litterers may react do not predict the intervention decision. We also find that the willingness to intervene is fundamentally different from the willingness to pay for a public awareness campaign that calls on others to intervene. These two decisions are not correlated and they differ in their predictors and their responsiveness to appeals or the provision of information.
As a member of a faculty of a university, I was asked each year to report on the number of conferences I had attended and the number of keynote addresses I had delivered and the number of research projects I had obtained funding for and the number of articles I had had published in refereed journals and the number of consultancies I had engaged in. I answered each such question by writing NIL in letters of a modest size and I returned each list of questions promptly to its sender and hoped I might be allowed to go on for a few more years teaching from February to November the eighty students who enrolled each year in my course and writing fiction from November to February.
Abstract: We study the coevolution of behavior, expressed beliefs and social preferences in a population engaging in pairwise cooperation through either a snowdrift game or a prisoner’s dilemma. We assume that individuals derive utility partially through the material payoff from cooperation, and partially through “norm-utility” (see e.g. Gavrilets et al, Phil Trans B, 2024), which captures the psychological benefit of adherence to or deviation from perceived social norms. In particular we explore the differences between the evolution of a population adhering to descriptive norms (derived from beliefs about what others actually do) as opposed to injunctive norms (derived from beliefs about what others should do). We identify a surprising trade-off, in which descriptive norms tend on average to promote greater cultural tightness (i.e. the existence of norms which are widely understood and adhered to), while at the same time disrupting social cohesion. Specifically, a shift from injunctive to descriptive norms facilitates evolutionary branching, in which the population splits into two groups who not only behave differently, but also possess different social preferences.
Abstract: There is by now a considerable amount of theoretical research on the evolutionary foundations of human motivation, or preferences. In the first part of this talk I will present a result which points to a particular class of preferences, which under general conditions combines a certain kind of pro-social concern (a Kantian moral concern) with a certain kind of anti-social concern (spite). The result further suggests that these two motivations go hand in hand. In the second part I will present experimental evidence based on simple designs that enable disentangling these motivations.