It might seem obvious to a secular person raised in a WEIRD culture (Henrich et al., Reference Henrich, Heine and Norenzayan2010) that a Mentawai water spirit does not actually punish people for failing to share food (Singh, target article). Singh argues these beliefs and practices culturally evolve within groups because they make intuitive sense to individuals, regardless of causal relations or objective outcomes. Yet, the article would benefit from turning this subjective selection lens on itself: What are examples of WEIRD practices that do not accomplish their purported goals but nevertheless persist because they are intuitive? Where do these intuitions come from? And which group members have them?

Western punitive justice systems offer such an example. Many scholars argue that third-party punishment – the imposition of costs on selfish actors by unaffected, often institutionalized, parties – is essential to the emergence and stability of large-scale human cooperation (Boyd, Gintis, & Bowles, Reference Boyd, Gintis and Bowles2010; Fehr & Gächter, Reference Fehr and Gächter2002, Reference Fehr and Fischbacher2004; Henrich et al., Reference Henrich, McElreath, Barr, Ensminger, Barrett, Bolyanatz and Ziker2006; Mathew & Boyd, Reference Mathew and Boyd2011). However, outside of theoretical models and laboratory experiments, evidence for this claim is weak. Meta-analyses consistently find that harsher penalties, such as “three strikes” laws, or even the death penalty, do not reliably reduce crime (Chalfin & McCrary, Reference Chalfin and McCrary2017; National Research Council, Nagin, & Pepper, Reference Nagin and Pepper2012; Nagin, Reference Nagin2013), and in the United States, 82% of people released from prison are rearrested within a decade (Antenangeli & Durose, Reference Durose and Antenangeli2021). Nonpunitive approaches such as restorative justice conferences – structured dialogues among offenders, victims, and community members – reduce recidivism relative to conventional criminal-justice processing (Sherman et al., Reference Sherman, Strang, Mayo-Wilson, Woods and Ariel2015; Strang et al., Reference Strang, Sherman, Mayo-Wilson, Woods and Ariel2013), yet do not receive the same level of support in WEIRD countries.

How do we reconcile the widespread emergence and persistence of harsh punitive institutions in WEIRD societies with their limited efficacy? Singh’s framework provides an answer: punishment feels like it ought to improve cooperation to WEIRD people, which is consistent with our own data. In our experiments, we document economic game conditions under which the inclusion of punishment is actively detrimental to cooperation (Alam & Rai, Reference Alam and Rai2025). These games were designed to mimic real-world conditions in which punishment can sometimes be profitable for punishers – for example, by extracting fines or resources from those they sanction (Sheth, Reference Sheth2016). Critically, we also find that people will choose to play in economic game conditions that incentivize severe punishment even as it reduces their own payouts (Alam & Rai, Reference Alam and Rai2025). When asked about their expectations, most believed that the presence of punishers would increase cooperation, and that paying punishers to engage in severe punishment would be most effective (Figure 1).

Figure 1.

The left panel shows lay intuitions about incentives and punishment, while the right panel shows the proportion of participants who choose to play in different game conditions and the expected values of their choices. From Alam and Rai (Reference Alam and Rai2025).

Why do WEIRD participants subjectively feel like punishment should increase cooperation? According to Singh, subjective selection operates by targeting proximate psychological goals. In the context of punishment, retribution may be the proximate goal of punishment rather than deterrence (Carlsmith, Reference Carlsmith2006). Yet, it is unclear whether these proximate retributive sentiments culturally evolved for the ultimate goal of inducing cooperation, or whether Western punishment practices may have culturally evolved for exploitation and domination. The earliest written legal codes, such as the laws of Ur-Nammu and Hammurabi in Mesopotamia, reserve the harshest penalties for women and slaves who harmed free men (Dunbar, Reference Dunbar2018). In the antebellum South, slave patrols maintained racial hierarchies by violently punishing enslaved people who attempted to escape or resist (Potter, Reference Potter2013). After abolition, these practices were transformed into convict leasing, in which Black Americans were criminalized en masse and forced into labor camps (Muhammad, Reference Muhammad2022). Echoes of this dominance function persist: for example, Black defendants are significantly more likely to receive a death penalty, especially if they are convicted of murdering a White victim (Sidanius & Pratto, Reference Sidanius and Pratto1999).

Thus, the modern belief in punishment as necessary for cooperation may reflect a process of subjective selection even as punishment itself culturally evolved for hierarchical purposes. This would suggest that explaining subjective selection in terms of intuition is incomplete without also taking into account the socio-historical conditions that make different explanations more cognitively accessible and socially motivated. In the context of western punishment, one possible explanation is that once Enlightenment ideals of liberty and equality became the preferred social-relational model for organizing legal relations in WEIRD societies, explaining that punishment served hierarchy became less cognitively accessible and carried reputational costs (Rai & Fiske, Reference Rai and Fiske2011; Singh & Hoffman, 2025).

Singh’s framework also invites a deeper question that remains unaddressed: intuitive to whom? The intuitive appeal of punitive justice is unevenly distributed (Johnson, Reference Johnson2008), and this may be due to poor perspective-taking. In one experiment, we asked participants to imagine themselves in different roles in a dictator game (Alam & Rai, Reference Alam and Rai2025). Those who imagined themselves as dictators, who might be subject to punishment, were more skeptical of punishers and their motives than players who were not subject to punishment. Dictators were also more likely to believe in the importance of never punishing fair behavior, while other players focused on the importance of always punishing selfish behavior. The same punitive system thus felt different depending on where people imagined themselves within it – either as potential targets or bystanders beyond its reach. This perspective gap may help explain why White voters often support “tough on crime” policies (Johnson, Reference Johnson2008) without considering what those policies signal to the minority communities they disproportionately target, where trust in law enforcement is low. Crucially, the people for whom it feels intuitive that punishment fosters cooperative are also those with disproportionate cultural and political influence – allowing the intuition to shape institutions and endure despite its empirical weakness. Singh’s account, therefore, could be strengthened by clarifying whose intuitions shape which cultural practices persist. However, doing so might force the theory to integrate subjective selection with the kinds of prestige, dominance, and conformity-based cultural transmission pathways that Singh seeks to distinguish subjective selection from (Henrich & Gil-White, Reference Henrich and Gil-White2001).