Conflicts are ubiquitous between individuals within communities and organizations as well as between groups, including societies and nations. Effective conflict resolution processes are an essential component of individual well-being, group functioning, and global security. Leadership is often a key component of conflict management and groups often “live or die” based on their ability to effectively settle variable conflict scenarios. The study of conflict resolution is a multidisciplinary pursuit with profound implications within academia as well as in politics, management studies, and industry. The evolutionary human sciences, including psychology and anthropology, have suggested conflict resolution and management is shaped by psychological heuristics and social and cultural variability. Despite much discourse and theorizing there is limited empirical data available on conflict resolution processes across the full range of human cultural and social diversity. This is a major limitation for conflict studies; in order develop broadly generalizable theories and effective policy recommendations a more detailed understanding of conflict resolution across human cultures and contexts is required. The current study draws on a novel cross-cultural database developed from primary ethnographic documents from representative sample of cultural diversity to identify variation in evidence for conflict resolution across cultures and contexts and to identity the traits of individuals who resolve conflicts, and the individual costs and benefits of conflict resolution across human societies. There was robust evidence of within-group conflict resolution which did not substantially vary across cultures with different subsistence strategies or group context, with two exceptions – leaders of military groups were involved in greater between-group conflicts and leaders of religious groups were involved in greater within-group conflicts. Exploratory analyses revealed that other group-beneficial functions were associated with conflict resolution including group representation, counsel, protection, and punishment, as well as the qualities of interpersonal skills and fairness. Whereas recipients of conflict resolution services received only the direct benefit of conflict resolution, providers of conflict resolution services received a wider range of indirect benefits including resources, mating benefits, and social services; no costs were strong predictors of conflict resolution for either recipients or providers. Lastly, there was evidence kin groups exhibited greater conflict resolution processes than political groups. These results provide a rare view of the correlates of conflict resolution processes and their universality across cultures and diversity across social contexts.


Humans and human societies are exceptional in our abilities to sustain large-scale, complex cooperation (Henrich & Henrich, 2007), including between relatives, strangers, residential communities, and political groups. Nature, including human sociality, is clearly not “red in tooth and claw” as Tennyson (1850) famously composed and diverse non-human species demonstrate and depend on substantial cooperation (e.g., Allee, 1951). Such cooperative relationships, however, are vulnerable and can quickly decay in the face of escalating and unresolved conflicts, particularly between autonomous human populations, such as extended kin groups and nation states. Group living necessarily creates ample opportunity for inter-individual conflict and imposes costs on individuals. Therefore, conflict resolution processes are expected to be an essential and universal feature of human groups (Boehm, 1982; Brown, 1991; Garfield, Syme, & Hagen, 2020). The underlying correlates of individuals who function to resolve conflicts within and between groups and the specific cultural and group contexts that present greater conflict resolution demands have been widely discussed across scientific disciplines, but empirical evidence from a representative sample of human social and cultural diversity has remained unavailable. Through novel exploratory comparative analyses the present study aims to fill these empirical gaps.

Sociality and demands of conflict resolution

Self-interested individuals in close proximity can be in conflict over access to subsistence or material resources, mating opportunities, and preferential territory, for example (Morris, Bastock, & Moynihan, 1954; Parker, 2006; Ross, 1983; Smith, 1985). Long-term co-residents and close kin can develop additional conflicts within a social or organizational structure including over position in a status hierarchy or social network and in economic and reciprocal exchange (Berger, Rosenholtz, & Zelditch, 1980; Boulding, 1957; Hames, 2015; Parker, Royle, & Hartley, 2002). Moreover, groups of individuals across all levels of social organization (e.g., kin, residential, economic, or political levels) can develop various inter-group conflicts (Glowacki, Wilson, & Wrangham, 2017; Roscoe, 2009; Wrangham & Glowacki, 2012). The evolution and maintenance of group living therefore faces significant challenges stemming from, in part, the numerous opportunities for conflicts within and between groups – a point further exemplified given that most species, including about 70% of all mammals, do not live in groups potentially as a consequence of these challenges (Wilson & Reeder, 2005).

For a minority of extant mammals including humans and most primates, group living is obligatory (Hrdy, 2006). Group living then must have in some contexts among some species, offered individuals a net fitness benefit over evolutionary history (Alexander, 1974). Putative benefits include reduced individual risks from heard formation, coordination of individuals to accomplish highly profitable yet difficult goals, and increased abilities to monitor and patrol territory (Bickerton & Szathmáry, 2011; Smith, Swanson, Reed, & Holekamp, 2012; Willems & van Schaik, 2017). In the context group living, co-residents who develop and maintain effective conflict resolution mechanisms and reliably solve inter-individual conflicts can benefit both at the individual-level via reduced aggression (direct and indirect) and increased cooperation (Alexander, 1974; Chapman & Valenta, 2015; Hess & Hagen, 2006) as well as at the group-level via cultural evolutionary group selection processes (Richerson et al., 2016; Smith, 2020) and gains from international cooperation (Snidal, 1991; Uitto & Duda, 2002).

Conflict resolution mechanisms are common across gregarious nonhuman species and include reconciliation behaviors between individuals following an aggressive conflict as well as intervening behaviors by a third-party, often a high-ranking or influential group member (Aureli, Cords, & van Schaik, 2002; Aureli, Waal, & Waal, 2000; Wilson, 1980). Collectivist philosophies and mantras, i.e., “united we stand, divided we fall,” are a common feature of national identity and political philosophies espoused by leaders to their constituents (Jacobs, Aesop, & Hejduk, 1991). Theoretical models also suggest cooperative dynamics can emerge in heterogeneous, conflict-prone groups when, for example, dominant leaders provide policing and conflict resolution services (Frank, 1996; Mesterton-Gibbons, Gavrilets, Gravner, & Akçay, 2011).

Conflict resolution in comparative perspective

Smith et al. (2016) provide an empirical cross-species and cross-cultural comparison of within-group conflict resolution, contrasting nonhuman animal populations with small-scale human populations. Across their sample, within-group conflict resolution functions were generally not widely distributed, i.e., a small number of adults had greater influence in conflict resolution in both nonhuman and small-scale human societies. This strongly implicates the role of individual leaders, or individuals who maintain a disproportionate level of influence within a group (Kantner, 2010; von Rueden & van Vugt, 2015), in the resolution of within-group conflicts.

Smith et al. (2016) also found that among both nonhuman species and small-scale human societies leaders and followers generally equally benefit from within-group conflict resolution (though the specifics of ‘relative benefits’ were not coded for) and conflict resolution roles are more often achieved rather than ascribed, suggesting followers and group dynamics shape the qualities of leaders who resolve conflicts. In small-scale human societies, however, leaders who resolve within-group conflicts are more likely to have to coercive power (Smith et al., 2016). The Smith et al. (2016) sample is small (n = 16) and empirical measures based are ratings based on ethnographic and ethological expertise, however they provide a novel and useful framework for developing a broader view of conflict resolution among humans and in comparative evolutionary perspective.

Among human populations, and smaller-scale societies in particular, social norms, cultural institutions, and supernatural beliefs can prevent and mediate many potential or actual instances of inter-individual conflict without direct resolution efforts of a third party (Basedow, 1925; Glowacki, 2020). For example, the ekila and poison oracle practices among Central African foragers (Lewis, 2008) and other supernatural mediation and divination rituals can function to resolve conflicts without much decision-making autonomy or influence from a third-party community member. Nonetheless, concerted dispute settlement by a third party is universal among human groups in at least some contexts (Brown, 1991; Garfield et al., 2020). Comparative and field working anthropologists are uniquely situated to provide in-depth and nuanced views of inter-personal and inter-group conflict resolution mechanisms Such perspectives are necessary to fully understand the cultural diversity and universality of human conflict resolution. Given the broad phylogenetic distribution of conflict resolution behaviors and the universality of conflict resolution processes among human populations, there are likely to be evolved psychological underpinnings to human behavioral dynamics associated with conflict resolution strategies, as well as reliable facultative responses to ecological conditions, and culturally evolved systems exhibiting variability and convergence.

Culture and conflict resolution

Early ethnographers extensively documented cultural variation in typologies of conflicts such as feuds, warfare, and fighting, but cohesive theories grappling with cultural variation were developed much later following influential sociological theories (e.g., Durkheim, Marx, Weber) (for review see Levine, 1961; Sluka, 1992). British sociocultural and interpretive anthropologists, e.g. Gluckman (1956) and Turner (1970), suggested conflicts within larger social units reinforce and contribute to the cohesiveness of larger social structures. Structural anthropologists, however, (e.g., Beals & Siegel, 1966) suggested conflict is a deleterious consequence of various “pressures” exerted on social systems, and often contrasted intrafamilial, intracommunity, intercommunity, and intercultural conflicts.

More recently, evolutionary anthropologists have emphasized the important role of leaders, senior kin, and other influential community members as the managers of conflicts withing and between families and kin groups and have suggested such conflict resolution processes underlay much of human psychological, biological, and cultural evolution (Boehm, 2013; Garfield, Hubbard, & Hagen, 2019; Glowacki & von Rueden, 2015; Hames, 2015). Drawing on the ethnography of smaller-scale, politically autonomous populations and empirical field data Glowacki & von Rueden (2015) frame within-group conflict resolution as one type of collective action problem populations must overcome. They suggest effective leadership emerges and is selected for as a critical component in the resolution of collective action problems within an institutional system and individuals with wider social networks, greater knowledge, and physical fromidability will be more likely to effectively resolve inter-individual conflicts. Nonetheless, most of the evolutionary anthropological literature on conflict has focused on evolutionary underpinnings of warfare and between group conflict (see Glowacki et al., 2017; Sluka, 1992).

Study aims

Based on the current theoretical and empirical literature, outstanding questions on the nature of human conflict resolution include: 1) what are the underlying behavioral, personality, or other characteristics that characterize leaders who resolve conflicts across diverse cultural and social contexts?; 2) what are the specific costs or benefits individuals incur from conflict resolution processes?; and 3) how widespread is coercive authority or prosociality of leaders who solve disputes, across diverse societies? Lastly, 4) does group context or cultural typologies predict variation in human conflict resolution? The present study aims to fill these empirical gaps in the human conflict resolution literature through a detailed exploratory analysis of a novel cross-cultural database constructed from the ethnographic record.

Garfield et al. (2020) reported that resolving conflicts was the most cross-culturally frequent leadership function and among the most frequent of 109 leadership dimensions in a large ethnographic sample, documented in 78% of cultures and in 14% of all text records. Given the frequency of conflict resolution in the ethnographic record of leadership and important outstanding questions, further analyses are warranted.

Whereas Garfield et al. (2020) provide a broad exploratory analyses on the universality and variability of leadership dimensions, the current study focuses in detail on their Resolve conflict measure to descriptively analyze the presence of ethnographic evidence for conflict resolution processes across subsistence types (e.g., hunter-gatherers, pastoralists, or agriculturalits), and group context (e.g., residential, kin, or political groups), and to identify if this evidence is biased towards conflicts within or between groups. Text-analysis will be used to provide insight into the semantic content of the ethnographic record associated with conflict resolution. The leadership dimensions (qualities and functions of leaders) predictive of evidence for conflict resolution will identity the behavioral and phenotypic correlates of individuals who resolve conflicts and of the cultural models of conflict resolution across human societies. Lastly, culture and group-level measures will be used to predict evidence for conflict resolution to identify how conflict resolution may covary with cultural and social contexts.


Cross-cultural database

This study uses data from the leadership data package (zhgarfield & grasshoppermouse, 2019), a novel cross-cultural database originally constructed by Garfield et al. (2019) and expanded by Garfield et al. (2020), designed to capture a wide range of ethnographic content related to leadership from the 60-cultural Probability Sample Files (Naroll, 1967) of the electronic Human Relations Area Files (eHRAF). The eHRAF is an electronic text database of primary ethnographic documents which can be queried using a thorough subject code system (Outline of Cultural Materials or OCM codes) and/or by keyword, at the paragraph level. The leadership data package is based on 1,212 ethnographic paragraphs (termed text records) extracted from the eHRAF using a broad search strategy targeting general descriptions of leadership. It includes two data frames of researcher-coded measures, one of which includes variables on evidence for 109 dimensions of leader qualities and functions, including conflict resolution (their Resolve conflict variable), as well as measures of costs and benefits for leaders and followers. A researcher-coded measure of group context (see Table 1), the context of leader functions (i.e., within-group, between-group, or both), and culture and document-level metadata (e.g., subsistence strategy, year of publication, author) are also provided. The current study uses this database of 1,212 text records, the associated 109 leadership dimensions, and culture and document-level measures in novel exploratory, comparative analyses on conflict resolution. For additional details on the construction of the database see Garfield et al. (2019) and Garfield et al. (2020).

Data analysis

The first goal of the current study was to empirically describe the ethnographic record of human conflict resolution through text analyses. Ethnography-based data are fundamentally rooted in bodies of text authored by ethnographers and other professional writers. The semantic content of these documents can be analyzed using text analytic methods to supplement analyses from researcher-coded variables. A document-term matrix was developed of the corpus of text records and elasticnet logistic regression from the glmnet package (Friedman, Hastie, & Tibshirani, 2010) was used to identify the terms that predict evidence for conflict resolution (see Results for additional details). Elasticnet regression models are penalized regression models that are effective when the number of predictors is large relative to the number of observations (\(n<<p\)). Elasticnet regression models in the current study used the “lasso” penalty (\(\alpha=1\)), which will often set many coefficients to 0, thereby identifying the most important predictors among all covariates.

Building on results from Garfield et al. (2020), the second goal was to represent and describe the distribution of evidence for conflict resolution across cultures with distinct subsistence strategies and across distinct group contexts. That is, when there is evidence leaders resolve conflicts, is it biased to a particular cultural or social context? Also, the context of conflict resolution functions, as related to conflicts within or between groups, are compared across these categories. The third goal was to identify which leadership dimensions predicted evidence for conflict resolution. Two elasticnet logistic regression models were used to identify A) the qualities and functions of leaders who resolve conflicts and B) the cost and benefits for leaders and followers associated with conflict resolution processes. The fourth goal, also expanding on findings from Garfield et al. (2020), was to more closely examine variation in conflict resolution by cultural and social variation, namely by subsistence strategy (see Figure 1) and group context (see Table 1), using logistic mixed effects models and the lme4 package (Bates, Mächler, Bolker, & Walker, 2015).

All analyses were conducted with R version 4.0.2 (2020-06-22).


The leadership data package sample comprises 1212 text records from 321 documents describing 59 cultures. Detailed descriptive results of ethnographic paragraphs in the sample are given in Garfield et al. (2020), including discussing of potential bias by ethnographer gender, year of publication, total pages of ethnography per culture, etc. (overall, potential biases along these lines were deemed to be negligible). The geographic distribution of the culture sample is reproduced here in Figure 1.