Conspecific violence has been pervasive throughout evolutionary history. The current research tested the hypotheses that individuals implicitly categorize combative contexts (i.e., play fighting, status contests, warfare, and anti-exploitative violence) and use the associated contextual information to guide expectations of combative tactics. Using U.S. and non-U.S. samples, Study 1 demonstrated consistent classification of combative contexts from scenarios for which little information was given and predictable shifts in the acceptability of combative tactics across contexts. Whereas severe tactics (e.g., eye-gouging) were acceptable in warfare and anti-exploitative violence, they were unacceptable in status contests and play fights. These results suggest the existence of implicit rules governing the contexts of combat. In Study 2, we explored the reputational consequences of violating these implicit rules. Results suggest that rule violators (e.g., those who use severe tactics in a status contest) are given less respect. These are the first studies to implicate specialized mechanisms for aggression that use contextual cues of violence to guide expectations and behavior.
KeywordsIntrasexual competition Social learning Combat Contextual violence Fighting behavior
What I’m asking about fighting is, is there something going on here that is the same as occurs in incest or kinship or anything else, namely that there are certain inherent rules of violence? Is there something “inherent” about the very structuring of fighting, and if so, then what is it? (Fox 1977:136)
Historical and forensic evidence reveals a persistent presence of violence throughout our evolutionary history (Keeley 1996; LeBlanc 2003; Walker and Bailey 2013). As such, natural selection has shaped our physiology (Covassin et al. 2012; Lassek and Gaulin 2009) and psychology (Fillingim et al. 2009; Sell et al. 2009a; 2009b) to manage the adaptive problems of aggressive conflict. Previous research has revealed context-specific mechanisms for violence, and in the current work, we argue that these mechanisms function as part of a combat network—psychological architecture that monitors cues available during violence and uses them to guide expectations and behavior. We begin our argument by discussing the context-dependent nature of violence along with the social learning required to integrate local information about combat behavior. Following this discussion, we present two studies that test the implications of the combat network, and we attempt to elucidate Fox’s (1977) query about the structure of violence.
Strategic Violence and the Contexts of Combat
Because violence undermines fitness, only strategic violence was maintained by selection (Pinker 2011). In a competitive environment that promotes strategic violence, the architecture for prohibiting others from encroaching on our fitness must be complex. To capture a glimpse of this complexity, consider how the brain assesses the formidability of intrasexual rivals to guide decisions of entering or fleeing from combat, calculates the relative costs and benefits of joining a raiding party, and concludes whether to punish or seek reparative strategies against exploiters (e.g., Mathew and Boyd 2011; Maynard Smith and Price 1973; Petersen et al. 2012). In short, a combat psychology must be prepared for violence in a variety of complex contexts. Accordingly, the intricacies of strategic violence have been elucidated by adopting a context-sensitive perspective (Goetz 2010; Griskevicius et al. 2009; Liddle et al. 2012), an approach that has highlighted specific circumstances in which aggressive behavior would have been adaptive. This approach has already led to the identification of functionally distinct combative contexts, such as intimate partner violence (Buss and Duntley 2011), warfare (Van Vugt 2009), and status contests (Wilson and Daly 1985). Here, we focus on four contexts in which individuals engage in violent behavior: play fighting, status contests, warfare, and anti-exploitative violence. Although these contexts are not a complete representation of the spectrum of strategic violence, we use them here because (1) they are the roots of a significant amount of violence that has occurred over human evolutionary history; (2) as argued below, each of these contexts uses violence for different purposes; and (3) previous research has used similar contexts of violence to describe the motivations for deploying aggressive behavior (Baumeister and Vohs 2004; Pinker 2011).
Rough-and-tumble play manifests in several nonhuman animal species (see Bekoff and Byers 1998 for a review) and in many cases resembles sparring bouts, wherein participants actively restrict aggressive behavior to minimize injury. From an evolutionary perspective, play fighting has been explained as a means to develop the motor behavior required for hunting or predator avoidance, or to develop and make assessments about one’s own and others’ ability to fight (Boulton and Smith 1995; Symons 1978). Despite a variety of evolutionary pathways in the development of play fighting, these bouts are a distinct type of agonistic behavior. Cross-culturally, boys engage in more play fights than girls, and commonly used tactics include chasing, kicking, hitting, and wrestling (Pellegrini and Smith 1998). From a young age, boys identify their own acts of play fighting as a friendly behavior that should not involve damaging strikes (Smith et al. 2004) and readily distinguish between play and serious fights (Nabuzoka and Smith 1999; Smith 2010).
Like play fighting, contests over status are commonly found among nonhuman animals (Maynard Smith and Price 1973) and, though difficult to account for empirically, are one of the most frequent forms of men’s intrasexual combat (Hilton et al. 2000). Status contests may have the purpose of negotiating dominance over territory, resources, or other males, or to access mates (Maynard Smith and Price 1973). Although the procedures of human status contests vary cross-culturally, there are notable consistencies. For example, contests are more likely to erupt between symmetrical opponents (individuals of similar status and formidability) since men of high status can ignore challenges from low-status men without risking loss of face (Gould 2003; Wilson and Daly 1985). When describing nonhuman animal contests, researchers have often noted a structured escalation (e.g., Clutton-Brock et al. 1979; Payne 1998), and even among humans, escalation is predictable. The conflict begins with a challenge (an insult or a threat), progresses to minimal physical conflict (posturing or pushing), transitions into violent combat (punches, kicks, and other fighting tactics used at full force), and ends when a combatant has submitted or has lost the capacity to continue the fight (Sell 2011). Although women sometimes engage in status contests, they more frequently use indirect aggression such as gossiping about individuals with whom they are in conflict, spreading malicious rumors, and breaking ties (Björkqvist et al. 1992). It is men who are more likely to use direct violence in their contests by engaging in fistfights (Cobbina et al. 2010), stick fights (Abbink 1999), and wrestling bouts (Llaurens et al. 2009).
Intergroup violence, team aggression, coalitional conflict, collective violence—although warfare has known several academic labels, it is understood as a conflict between predefined groups. Wherever traces of ancestral society are found, evidence of warfare is exists (Gat 2006; Keeley 1996). Composed of primarily male, close-knit squads, warfare is violence on a scale larger than any other form of combat, seen in disputes over scarce resources, the settling of group vendettas, and the introduction of new women into the group (Chagnon 1988; Potts and Hayden 2008). War is a conflict carried out by men: 97% of soldiers in the world’s standing armies are men, and 99% of military positions designated for combat are occupied by men (Goldstein 2001). Estimates reveal that up to 30% of ancestral men died from war (Keeley 1996) and that warfare has been a common occurrence (Divale 1972; Ember 1978). Historically, warfare has driven technological advancements, such as weapons to be used in attacking enemies (Potts and Hayden 2008). The persistence of warfare has also shaped a specialized coalitional psychology in men (Van Vugt 2009; Wrangham 1999). Research investigating coalitional psychology has found that young boys and adult men report a desire for affiliation with formidable allies (DeKay et al. 1998; Goetz et al. 2009), that intergroup threats enhance men’s but not women’s discrimination between ingroup and outgroup members (Yuki and Yokota 2009), and that men more readily cooperate with ingroup members in the context of intergroup competition (Van Vugt et al. 2007).
Anti-exploitative violence arises from the arms race between mechanisms designed to exploit others (Buss and Duntley 2008) and to counter such exploitation (Duntley and Shackelford 2012; Petersen et al. 2012). Exploitative violence, when precisely timed at a strategically targeted person, can minimize or eliminate costs to the aggressor (Gat 1999). From the viewpoint of the exploiter, violence can effectively eliminate rivals and obtain their resources (e.g., by ambushing opponents). Buss and Duntley (2008) identified mugging, sexual assault, stalking, and cuckoldry as potential areas in which individuals might seek to exploit others; hence, treachery, surprise, and asymmetrical violence are the tools of an exploiter. In contrast, from the perspective of the anti-exploiter, violence is used to defend oneself or others from being exploited by avoiding suspicious individuals (e.g., stranger anxiety), entering a fetal position to protect vital organs, and eliciting post-victimization behaviors that will reduce the likelihood of recurrence (e.g., seeking revenge or retribution; Duntley and Shackelford 2012). Similarly, Petersen et al. (2012) noted that punishment and efforts to repair relationships with exploiters are also incorporated into the arsenal of counter-exploitation. Although both men and women are capable of deploying exploitative strategies, men’s physiology and psychology are more readily capable of engaging in exploitative violence (see Sell et al. 2012 for a review). Moreover, because men are regularly the victims of other men’s violence (Daly and Wilson 1988), they are also likely to be frequent users of anti-exploitative tactics. Although exploitative strategies may have provided fitness advantages, they are often socially inexcusable (Duntley and Buss 2004). Moreover, anti-exploitative strategies—for example, homicide in the form of self-defense or in the prevention of rape—can be viewed as justified or even altruistic (Duntley and Buss 2004).
Although the contexts are functionally distinct, humans recurrently engaged in each of them. The ability to compete in each context was likely heavily dependent on phenotypic (e.g., formidability and health) and environmental (e.g., culture and experience) factors. Previous research has made it clear that human biology, particularly that of males, was shaped for competition in these contexts (Puts 2010; Sell et al. 2012), but what is less clear is how the mind has been shaped to acquire knowledge about these contexts.
Prepared Learning for Combat
The Current Research
Despite obvious cross-cultural differences in how violence is carried out, context-specific violence has the same ultimate functions—in other words, we can expect violent behavior to vary non-functionally across cultures (e.g., whether a society settles status disputes with wrestling bouts or boxing rounds) but anticipate universally accepted violent behavior that is contextually functional (e.g., the approval of violence that would maim a violent exploiter but not a play fight opponent). As Wood (2007) argues, the narratives of cultural violence are pliable but not endlessly changing. Given the highly strategic and context-specific nature of violence and the social (among other) ramifications of combat, humans erect underlying rules—sets of behavioral expectations, sanctions, and prohibitions that are implicitly wielded by combatants, monitored by witnesses, and constrained by context. In testing this hypothesis, we predicted that (1) combat is implicitly monitored and is classified from context (i.e., without explicit prompting to attend to any particular details of a violent scenario, humans encode the relevant information that allows them to determine the context), and (2) regulations on combat behavior (i.e., sanctions of violent tactics) are shaped to facilitate the function of the context-specific conflict. Specifically, combative tactics in play fights should be the most limited in severity; only violence that imposes minimal damage should be acceptable in a play fight. Compared with play fights, status contests should increase the acceptability of more severe tactics, but considering that the aim of status contests is not rival elimination, highly lethal tactics should not reach widespread acceptance. Moreover, compared with status contests, warfare should further increase the acceptability of severe tactics, including the acceptance of lethal tactics. Lastly, given the direct costs to victims of exploitative violence, the general approval of punishing cheaters, and tolerance for excess force used in self-defense, anti-exploitative violence should be least constrained by the severity of violence, achieving large-scale acceptability despite costs to the exploiter. Although support for these predictions is not dependent on sex differences, the overrepresentation of men in combat suggests that their assessments might differ from women’s. In accordance, we probe for sex differences and discuss the relevant results.
A total of 237 (74 men, 163 women, mean age = 19.61) students from a Southern California university were recruited for participation in the study. Participants self-reported their ethnicities as African American (3%), Asian American (32.1%), Caucasian (32.1%), Hispanic (24.1%), or other (8.7%).
Also, we included standard demographic questions (age, sex, country, etc.) and a self-report measure of time spent explicitly discussing appropriate combat behavior.
Participants completed a repeated measures online experiment in which they were presented with a randomly assigned scenario. Upon reading the scenario, participants used a 10-point scale to rate the 22 combative tactics in terms of their perceived acceptability. Specifically, participants were asked, “Which of the following actions are acceptable for either man to do to the other?” After rating the 22 tactics, participants were presented with a multiple-choice question that asked them to identify the context that best reflects the type of conflict that was occurring in the scenario (play fight, status contest, anti-exploitative [represented as self-defense], warfare, and combat sports [a distractor]). Afterward, participants were randomly assigned another scenario and asked to repeat the measurements. After participants responded to all four scenarios, demographic data were collected.
Prediction 1a: Do Participants Accurately Categorize Contexts?
Binomial tests of scenario classifications
Study 1a n = 237
Study 1b n = 91
Are There Sex Differences in Classification Accuracy?
No. Although examination of percentages suggests that men are slightly more accurate than women at classifying contexts such as status contests (82.43% vs. 74.23%) and warfare (70.27% vs. 66.87%), chi-square tests indicate that these differences are not statistically significant (see the ESM for the complete results).
Prediction 2a: Do Anticipatory Rules of Combat Align with the Function of Combative Contexts?
However, a more nuanced understanding of the results is provided by appreciating the interaction between context and tactics. Both play fights and status contests feature similar acceptance ratings for the least severe tactics while having a similar decrease in acceptance for the most severe tactics. However, play fights and status contests most differ from one another when considering tactics of mid-severity (e.g., punching the face and kicking the mid-body). It is among mid-severity tactics that status contests elicit the most notable increases in acceptability, indicating that expectations in this context are neither to eliminate an opponent nor for a docile bout. Play fights on the other hand only allow for low-severity tactics, representing the friendly nature of the bouts.
When warfare is compared with status contests, a widespread increase in acceptability is noted. Though the differences between the two contexts are least noticeable in low-severity tactics, warfare prominently raises the acceptance ratings from mid-severity tactics through the most severe tactics. The increase in acceptability of more-damaging tactics is indicative of warfare’s potential to maim or eliminate opponents.
Anti-exploitative violence demonstrates the highest rates of acceptability across all tactics. Although acceptability of high-severity tactics is increased in anti-exploitative violence, it can be distinguished from warfare in terms of low- and mid-severity tactics.
Are There Sex Differences in Tactic Acceptance across Contexts?
Yes. Sex had a significant main effect on ratings, F1, 225 = 16.40, p < 0.001, ηp2 = 0.068, and it interacted with context, F3, 675 = 7.46, p < 0.001, ηp2 = 0.032, in addition to ineracting with tactics, F21, 4,725 = 2.68, p < 0.001, ηp2 = 0.012. A three-way interaction between context, tactics, and sex was also documented, F63, 14,175 = 2.11, p < 0.001, ηp2 = 0.009. Visual displays for context and tactics, as well as results from the vertical pairwise comparisons, are provided in the ESM, but to facilitate a concise interpretation of the influence of sex, we review here the interaction of context and sex. Although men (M ± SE = 1.937 ± 0.104) and women (M ± SE = 1.936 ± 0.068) did not differ in their ratings of tactics during a play fight, p = 0.997, men (M ± SE = 3.463 ± 0.175) were generally more accepting of the tactics than women during a status contest (M ± SE = 2.945 ± 0.116), p = 0.015. Additionally, men (M ± SE = 7.325 ± 0.299) were more accepting of tactics than women (M ± SE = 6.285 ± 0.197) during anti-exploitative violence, p = 0.004. However, the most apparent sex difference came in men’s willingness to accept tactics during warfare (M ± SE = 6.131 ± 0.294) in comparison with women (M ± SE = 4.612 ± 0.194), p = 0.001.
To determine if hypotheses 1a and 2a are supported outside the United States, Study 1b recruited participants using Amazon’s Mechanical Turk system (MTurk). This sample yielded an initial 183 cases; however, because of dropout (n = 30), missing data (n = 2), duplicate cases (n = 17), monotonic response patterns (n = 19), and other data anomalies (n = 24), a total of 92 cases were excluded from analysis. The final sample of 91 (60 men, 31 women, mean age = 28.41) was used for analysis. The non-U.S. sample was composed of individuals from India (82.4%), Australia (3.3%), Canada (3.3%), Croatia (1.1%), Hungary (1.1%), Pakistan (2.2%), the Philippines (2.2%), the United Arab Emirates (1.1%), Italy (1.1%), and the United Kingdom (2.2%).
Materials and Procedures
The same scenarios and questionnaires used in Study 1a were used in Study 1b. Participants were made aware of the study through the MTurk system and completed the experiment online using Qualtrics.
Prediction 1b: Do Participants Accurately Categorize Contexts?
Yes. Results from binomial tests (shown in Table 1) suggest that when participants were asked to classify the scenarios, they did so with a significant degree of accuracy across the four contexts.
Are There Observed Sex Differences in Classification Accuracy?
No. Although women demonstrated greater accuracy in status contests (83.87% vs. 71.67%) and anti-exploitative violence (74.19% vs. 63.33%), these differences were not statistically significant. The largest discrepancy in accuracy came in the context of warfare, for which men were more accurate than women (66.67% vs. 51.61%), but again, this difference was not significant (see the ESM for complete results).
Prediction 2b: Do Anticipatory Rules of Violent Behavior Align with the Function of Combat Context?
Yes. Analysis of the non-U.S. sample revealed a main effect of context F3,249 = 10.45, p < 0.001, ηp2 = 0.112, and a main effect of tactic acceptability, F21, 1,743 = 2.44, p < 0.001, ηp2 = 0.029. Furthermore, there was also a significant interaction between context and tactic acceptability, F63, 5,229 = 1.89, p < 0.001, ηp2 = 0.022. Analysis of the covariate, time spent explicitly learning about and discussing acceptable combat behavior, revealed that it influenced participants’ responses, F1,83 = 13.22, p < 0.001, ηp2 = 0.137; estimated means from the analysis were adjusted for this effect.
Although the range of acceptability ratings for the non-U.S. sample is more constricted than that of the undergraduates, similar patterns also manifest with regard to the interaction of context and tactics. Both play fights and status contests elicit a similar decrease in acceptability ratings for high-severity tactics and an increase in ratings for the least severe tactics. Again, status contest is most differentiated from play fight in terms of mid-severity tactics (e.g., punching or elbowing the face).
When warfare is compared with status contests, a sustained increase in acceptability is noted across tactics. Yet for warfare the most marked increase in ratings comes for mid- through high-severity tactics. Except for a few of the high-severity tactics (e.g., striking an unconscious opponent or using a weapon), the profile for anti-exploitative violence obtains the most acceptability. In terms of high-severity tactics, warfare and anti-exploitative violence are nearly indistinguisable. However, as with the sample of undergraduates, anti-exploitative violence is most distinct from warfare in terms of low- and mid-severity tactics.
Are There Sex Differences in Tactic Acceptance across Contexts?
No. Analysis of the main effect of sex revealed that men and women did not respond differently, F1,83 = 0.51, p = 0.479, ηp2 = 0.006. Moreover, sex did not interact with context, F3, 249 = 1.50, p = 0.214, ηp2 = 0.018, nor was there an interaction of sex with tactics, F21, 1,743 = 1.54, p = 0.06, ηp2 = 0.029. Lastly, there was no three-way interaction among context, tactic, and sex, F63, 5,229 = 0.70, p = 0.963, ηp2 = 0.018.
Study 1 demonstrated initial support for a combat network that monitors conflicts and guides the contextual acceptability of a tactic. Participants readily identified contexts when given scenarios with limited information about the onset of a confrontation; both American undergraduates and the older, non-U.S. sample had high rates of agreement on how scenarios should be classified. Moreover, participants predictably shifted their assessments of combative behavior across contexts.
Although the scenarios allowed participants to categorize conflicts with high reliability, flaws in their construction are evident. For example, the play fight scenario explicitly stated that one individual pushed another, and this in turn might partially explain increases in the acceptability of pushing for this context. Also, the warfare scenario stated that previous deaths had occurred due to the conflict between the groups, and this may have driven acceptability ratings for high-severity tactics. Because violence is not typically accompanied by rumination on what tactics should be employed, it is possible that the materials (scenarios and scales) did not fully activate the corresponding psychology or accurately capture the application of implicit rules. Another concern arises from the restricted range in responses demonstrated by participants. This is most noted in the reluctance of participants to endorse the highest rates of tactic acceptability even during warfare and anti-exploitative violence (see mean context and tactic figures in the ESM). Also of concern in the non-U.S. sample are the somewhat supportive responses for using severe tactics in a play fight—the lack of translation of the scenarios and measurements to suit non-English-speakers could be what underlies this finding. Likewise, this might explain why the pairwise comparisons were not as consistent in the non-U.S. sample as they were in the sample of undergraduates.
Notably, Study 1b did not replicate the sex differences found in Study 1a. Ultimately, future research and replications will determine the stability of sex differences across cultures; however, we can offer some limited speculations about what researchers should expect. One viewpoint suggests that we should anticipate stable sex differences across the contexts. Cross-culturally, men’s participation in these contexts vastly outweighs women’s participation, and differences in experience could generate different assessments. In this case, it is possible that methodological limitations (such as lack of scenario translation) failed to appropriately capture existing sex differences in the non-U.S. sample. Another, more likely perspective is that we should not always anticipate stable sex differences across cultures because combat is also relevant to women. Although women do not engage in nearly as much violence as men do, there could be important reasons why the sexes might have similar assessments across the contexts. The contexts of violence are not necessarily mutually exclusive; the implications of status, for example, are present across contexts. The monitoring of status consequences would have been of evolutionary significance for women as well as men, and as a result, women might adopt similar assessments of combat. Similarly, Sell et al. (2009a, 2010) found that women, like men, are able to accurately assess men’s formidability from body, face, and voice. It is not likely that women needed to assess men’s formidability because they were making decisions about whether or not they should fight a particular man but because formidability could be an indicator of other cues, such as resource holding potential, mate quality (Hill et al. 2013), or the likelihood a man would pose a danger to a woman. Thus, women may be expected to categorize contexts and assess tactics similarly to men. More specifically, women’s attention to violence should be also influenced by how commonplace violence is in their local environments. Women living in environments where violence is more frequently used (and is socially permissible) to negotiate conflicts should have similar interests and assessments to men’s. Despite these limitations, it is worthwhile to note the consistencies recorded between a WEIRD (Western, educated, industrialized, rich, and democratic; Henrich et al. 2010) sample and a sample that is much less WEIRD. These consistencies provide preliminary evidence for the universality of psychological mechanisms associated with monitoring the cues of violence, differentiating between the types of combat, and guiding expectations of violent behavior.
Of the reviewed contexts of combat, status contests serve as valuable forums for displaying indicators of formidability (i.e., resource holding potential) and mate quality (Maynard Smith and Harper 2003; Zahavi 1975). Hence, it is not surprising to find cases such as Sereer wrestlers who practice polygyny more often than non-wrestlers (Llaurens et al. 2009) and those of formidable Tsimane men who are frequent participants in extra-pair copulations (von Rueden et al. 2011). With regard to status contests, Sell (2011) referenced the existence of fair fights, contests that can give observers accurate information about the relative fighting ability of combatants (i.e., who the better fighter is), and that “cheap shots” obscure the ability to make accurate assessments.
Study 1 indicated that severe and lethal tactics were unacceptable in a status contest, setting a foundation for the rules of this context. If “cheap shots” inhibit the ability to judge who the better combatant is, then we should expect rule breakers to receive negative attributions. Moreover, if status contests are about proving who the better fighter is, then we should expect combatants who use self-handicapping tactics to receive positive attributions. In Study 2, we test these predictions with the attribution of respect.
A total of 234 participants (97 men, 137 women, mean age = 19.66) were recruited from undergraduate psychology classes at a Southern California university. Participants classified themselves as African American (2.1%), Asian (23.5%), Hispanic (32.9%), Caucasian (33.3%), or other (8.2%).
Measures and Procedure
Participants read the same status contest scenario used in Study 1. After reading the scenario, participants rated their attributions of respect (a facet of reputation) for six tactics on a 10-point scale. Three of the most severe and unacceptable tactics for a status contest (based on Study 1) were used: bringing allies into combat, striking the testicles, and using a lethal weapon. These tactics were also selected because they were judged to be behaviors that would inhibit the ability of third parties to assess who the better fighter is. Three novel tactics were used to represent user-imposed disadvantages: fighting using one arm only, fighting a rival despite the opponent bringing his allies into combat (i.e., fighting when outnumbered), and fighting an opponent who was significantly larger and more experienced in combat (i.e., fighting a formidable opponent). Participants completed the study online, where presentation order of the tactics was randomized.
Are There Reputational Costs for Violating the Implicit Rules of a Status Contest?
Yes. Results suggest a significant main effect of tactic, F5, 1,090 = 55.76, p < 0.001, ηp2 = 0.204. Results also show that the covariate, time spent explicitly learning about and discussing acceptable combat behavior, also had a notable main effect, F1, 218 = 8.62, p = 0.004, ηp2 = 0.038.
Are There Sex Differences?
Yes. The main effect of sex was found to influence respectability ratings, F1,218 = 22.26, p < 0.001, ηp2 = 0.093. Furthermore, there was also a significant interaction between tactics and sex, F5,1,090 = 6.05, p < 0.001, ηp2 = 0.027. Pairwise comparisons suggested that men generally attribute more respect than women do, but most prominently for the use of self-handicapping tactics (refer to the ESM for a visual display). Men (M ± SE = 7.79 ± 0.292) attributed more respect than women (M ± SE = 6.68 ± 0.245) when a combatant fought with one arm (p = 0.004), fought outnumbered (M ± SE = 7.30 ± 0.292 vs. 5.51 ± 0.245, p < 0.001), and fought a more formidable opponent (M ± SE = 6.64 ± 0.245 vs. 5.12 ± 0.206, p < 0.001). When analyzing tactics that violate acceptability in a status contest, men and women decrease respect ratings more similarly. No sex differences were evident for inviting allies (M ± SE = 2.64 ± 0.180 vs. 2.52 ± 0.152, p = 0.614) or using a weapon (M ± SE = 2.24 ± 0.180 vs. 2.04 ± 0.151, p = 0.409). However, men (M ± SE = 3.04 ± 0.232) did attribute more respect than women (M ± SE = 2.32 ± 0.195) to a combatant who would strike his opponent’s testicles (p = 0.02).
Study 2 facilitated an understanding of reputational consequences for violations of implicit rules in a status contest and reputational enhancements for demonstrating self-handicapping behaviors. Aside from violating expectations for a status contest, unacceptable (i.e., high-severity) tactics may also decrease respect because participants and witnesses anticipate that combatants will “fight fair” (Sell 2011), and these tactics might be perceived as unfair because they undermine the ability to honestly reveal the better combatant. Moreover, increased ratings for respect may also be driven by the bias to support underdogs (Vandello et al. 2007).
Study 2 has some of the same limitations as Study 1, including that the scenario and scales may not be fully activating the corresponding psychology and capturing the application of implicit rules. We find this limitation unlikely, however, as similar studies have had success activating specialized psychology through scenarios (e.g., Delton and Robertson 2012; Griskevicius et al. 2009), and rather than displaying random patterns, the results generally support the predictions. Another concern of both Studies 1 and 2 is that it is difficult to determine whether the findings are more dependent on the superficial details of each scenario or are indeed the outcome of understanding the inherent combative context. From an evolutionary perspective, we should anticipate that superficial traits would be incorporated into our understanding of contextual combat. For example, characteristics such as the status contest occurring in the presence of an attractive woman or a light-hearted insult being given in the play fight are superficial features of the scenarios; however, they serve to communicate underlying aspects of the contexts: third party witnesses during status contests, and the amiable nature that typifies play fights. Nevertheless, future research should consider generating novel scenarios that follow the general premise of each type of functionally distinct context that involves violent behavior. Another limitation of both studies concerns the question of why third parties are interested in assessing the acceptability of violence that does not directly concern them. A convincing case could be made if we consider that third parties might be better able to acquire information for coalition forming, assess formidability, and adjust welfare trade-off ratios by monitoring the violence of other individuals or groups. Further research should explore other possibilities that could have led to these abilities. An additional limitation could be that because Study 2 did not measure the same 22 tactics used in Study 1, we cannot directly assess whether low- or mid-severity tactics are given higher respect ratings. Also, since Study 2 was not preceded by non-contextual rankings of self-handicapping tactics, the tactics may not reflect self-handicapping. However, because self-handicapping in combat can be thought of as intentionally limiting one’s fighting potential in order to increase the probability of an opponent’s victory (see Boulton and Smith 1995), the tactics used in the study to represent self-handicapping appear to have good face validity.
After working with the Tory, Fox (1977) contemplated the structure of violence and its relation to human nature (see epigraph). It was evident that some animal species had an ordered process to their violence that avoided lethal tactics: for example, the initiation of dyadic tail-wrestling contests with agonistic displays among rattlesnakes (Almeida-Santos et al. 1999), the transition from tranquil swimming to fierce fighting among male squid when female reproductive pheromones are detected (Cummins et al. 2011), the cooperative wrestling contests of sierra dome spiders (Decarvalho et al. 2004), and the general restrictiveness of tactics of certain species when engaging in play fights (Bekoff and Byers 1998). Fox realized that the Tory could not explain the structure of their status contests despite the fact that the Tory men readily carried out the violence as if each person had read a long list of explicit rules to guide their behavior and expectations. Fox realized that there may be some underlying process of violence, which our results suggest are implicit rules of combat carried across the contexts of violence.
Future research on the topic should consider a military sample, which might elucidate whether individuals who have received explicit training in combat still uphold the assessments participants made in the current study. Another direction for future research is exploring different combative contexts (e.g., exploitative violence) and considering an extended list of violent tactics. Future research could also extend Study 2 by measuring respect ratings for low- and mid-severity tactics. This type of research could yield unique insight into the nature of human combat. For example, mechanisms that drive the social learning of combative contexts could partially explain the consumption of and demand for violent media. These mechanisms would have been designed to predispose humans to attend to the available information in the environment regarding violence in order to learn about the implicit rules and dangers of combat. Moreover, viewing combat as a part of human nature could suggest that it is universally understood: combat sports such as boxing, wrestling, or mixed martial arts could have an international appeal. Despite the fact that each sport has specific rules, humans might be predisposed to learn and understand them.
From an alien’s perspective, violence of all types might appear undifferentiated. This research demonstrates that our brains implicitly encode contextual information and can make qualitative classifications of violent contexts. Moreover, this study demonstrates that our expectations of violent behavior change according to context. Specifically, approval of the severity of violence depends on the function of the combative context. Violence that could kill is only supported in contexts that function to eliminate rivals or group threats, and violence that could cause serious harm is not supported in bouts of play fighting. When these expectations are not fulfilled, humans apply reputational punishments (i.e., loss of respect); when they are exceeded, human apply reputational enhancements (i.e., increased respect).
It is possible that everything we have demonstrated here is the result of social learning. Indeed, we argue that social learning is an important tool for learning local cultural information about violent tactics. In our analyses, we attempted to account for some of this variation by including self-reported time spent explicitly learning about and discussing acceptable combat behavior as a covariate. The analyses suggest that this variable inconsistently influenced responses, but that even when accounted for, the results still supported the a priori predictions. Culture is a product of evolved psychology, and despite the limitations of the current research, evidence is arising to suggest that the interaction of biological and cultural evolution has allowed us to erect a universal set of implicit combat rules.
The exploration of psychological mechanisms and social learning rules has enlightened our understanding of violence and, to an extent, we feel that the current research helps address part of Fox’s question. We understand that several contexts of violence are managed by psychological mechanisms, but as the current study suggests, the monitoring of these contexts is better understood within the framework of a network of combat mechanisms that implicitly guide and assess violent behavior—a system that imposes rules constrained by the strategic purposes of the violence. When coupled with social learning rules, a theoretical framework is constructed for individuals to monitor, learn, and execute context-sensitive violent behavior.
We value the thoughtful input given by Shiloh Betterley, Kayla Causey, Justin Lynn, and Elizabeth Pillsworth throughout this project.
- Baumeister, R. F., & Vohs, K. D. (2004). Four roots of evil. In A. Miller (Ed.), The social psychology of good and evil. New York: Guilford Press.Google Scholar
- Boulton, M. J., & Smith, P. K. (1995). The social nature of play fighting and play chasing: mechanisms and strategies underlying cooperation and compromise. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), Adapted mind: Evolutionary psychology and the generation of culture (pp. 429–444). New York: Oxford University Press.Google Scholar
- Covassin, T., Elbin, R. J., Harris, W., Parker, T., & Kontos, A. (2012). The role of age and sex in symptoms, neurocognitive performance, and postural stability in athletes after concussion. American Journal of Sports Medicine, 40(6), 1303–1312.Google Scholar
- Dahlberg, L. L., Toal, S. B., Swahn, M., & Behrens, C. B. (2005). Measuring violence-related attitudes, behaviors, and influences among youths: A compendium of assessment tools. Atlanta: Division of Violence Prevention, National Center or Injury Prevention and Control, Centers for Disease Control and Prevention.Google Scholar
- Daly, M., & Wilson, M. (1988). Homicide. New Jersey: Transaction.Google Scholar
- DeKay, W. T., Buss, D. M., & Stone, V. (1998). Coalitions, mates, and friends: toward an evolutionary psychology of relationship preferences. Unpuplished manuscript. Ann Arbor: University of Michigan.Google Scholar
- Divale, W. T. (1972). Systemic population control in the middle and upper paleolithic: inferences based on contemporary hunter-gatherers. Archaeology, 4, 222–243.Google Scholar
- Duntley, J. D., & Buss, D. M. (2004). The evolution of evil. In A. Miller (Ed.), The social psychology of good and evil. New York: Guilford Press.Google Scholar
- Fox, R. (1977). The inherent rules of violence. In P. Collett (Ed.), Social rules and social behaviour (pp. 132–149). Totowa: Rowman and Littlefield.Google Scholar
- Gat, A. (1999). The pattern of fighting in simple, small-scale, prestate societies. Journal of Anthropological Research, 55, 563–583.Google Scholar
- Gat, A. (2006). War in human civilization. New York: Oxford University Press.Google Scholar
- Goetz, A. T. (2010). The evolutionary psychology of violence. Psicothema, 22, 15–21.Google Scholar
- Goetz, A. T., Causey, K. B., Romero, G. A., & Black, A. (2009). Coalitional psychology in young children: Boys but not girls draw an imagined ally as taller and larger than themselves. Poster presented at the 21st annual meeting of the Human Behavior and Evolution Society, Fullerton, CA.Google Scholar
- Goldstein, J. S. (2001). War and gender: How gender shapes the war system and vice versa. Cambridge: Cambridge University Press.Google Scholar
- Gould, R. (2003). Collision of wills: How ambiguity about social rank breeds conflict. Chicago: University of Chicago Press.Google Scholar
- Keeley, L. H. (1996). War before civilization: The myth of the peaceful savage. New York: Oxford University Press.Google Scholar
- LeBlanc, S. A. (2003). Constant battles: Why we fight. New York: St. Martin’s Press.Google Scholar
- Maynard Smith, J., & Harper, D. (2003). Animal signals. New York: Oxford University Press.Google Scholar
- Pellegrini, A. D., & Smith, P. K. (1998). Physical activity play: the nature and function of a neglected aspect of play. Child Development, 69, 557–598.Google Scholar
- Pinker, S. (2011). The better angels of our nature: Why violence has declined. New York: Penguin Group.Google Scholar
- Potts, M., & Hayden, T. (2008). Sex and war: How biology explains warfare and terrorism and offers a path to a safer world. Dallas: Benbella Books, Inc.Google Scholar
- Smith, P. K. (2010). Children and play. Chichester: Wiley-Blackwell.Google Scholar
- Symons, D. (1978). Play and aggression: A study of rhesus monkeys. New York: Columbia University Press.Google Scholar