1Before the 1970s, primatologists typically portrayed non-human primate societies as male-dominated, characterized by violent competition among males for alpha status, with females relegated to secondary, passive roles. The work of C.R. Carpenter is often cited as emblematic of this perspective (Carpenter, 1964). This male-centric framing was subsequently criticized by Donna Haraway, a pioneering American scholar in gender studies, in her influential book Primate Visions: Gender, Race, and Nature in the World of Modern Science (1989). In it, Haraway argues that science is never truly neutral; rather, it is shaped by the cultural and societal biases of its time.
2According to Haraway, early primatologists projected their own gendered assumptions onto the primate societies they studied, effectively reinforcing and naturalizing the power hierarchies and gender roles that prevailed in various human societies. This critique illustrates a broader concern: that scientific practices, far from being objective, can contribute to the reproduction of social power structures. As a result, the study of power dynamics among non-human primates – and the biases embedded in such research – has become a key case study in epistemology.
3Haraway’s critique laid the groundwork for a broader framework that calls on modern societies to critically reassess the ideal of scientific neutrality. In Simians, Cyborgs, and Women: The Reinvention of Nature (1991), she further explores how science is intertwined with power systems and advocates alternative approaches to research – approaches that are more reflexive and better attuned to the unconscious biases that shape scientific discourse.
4This opening reflection raises core philosophical issues that remain important to consider for primatologists today. Research into primate power structures, whether intentionally or not, often intersects with broader societal debates about gender asymmetries and the presumed objectivity of science. These themes remain acutely relevant in contemporary discourse, underscoring the importance of critical, interdisciplinary engagement in both the production and interpretation of scientific knowledge.
5Whether influenced by gender studies, by the increasing feminization of the discipline, or simply by the growth and diversification of primate field studies, a second generation of primatology studies is indeed drawing a richer and more nuanced picture of sex roles in primate societies (Strier, 1994). Studies focusing on females revealed that they carefully chose their mates or actively solicited multiple mates (Keddy-Hector, 1992; Small, 1995), that they often formed the 'social cement' of primate groups (e.g., Silk, 2007) and that, like males, they competed fiercely over resources, rank or mates (Emery Thompson et al., 2008; Clutton-Brock and Huchard, 2013). In the 1980’s, Alison Jolly’s work on Malagasy ring-tailed lemurs reported that societies can be dominated by females (Jolly, 1984). In the 1990’s came the even more ground-breaking realization that some of our closest relatives, the bonobos, also live in female-dominated societies (Parish et al., 2000). In some South American monkeys like tamarins, there is no clear dominance relationship between males and females and it is males, not females, who provide the most parental care (e.g. Goldizen, 1989).
6Meanwhile, the parallel development of the theory of sexual conflict in evolutionary biology led to the formalization of male-female struggles over sexual and social control, and the broader realization that the interests of males and females often diverge, sometimes causing severe harm to individuals of one sex, or to their Darwinian fitness (Parker, 2006). This development broke with the traditional picture of sexual selection whereby males compete together and females choose a mate among the winners, which depicted more harmonious and complementary relationships between the sexes. This opened up new perspectives in the study of male-female sexual and social relationships, which revealed, in particular, how primate males attempt to control female sexuality, sometimes forcefully, and the diverse strategies that females develop to counter their attempts (Davidian et al., 2022).
7This broad and concise presentation of three important academic blocks in primatology, philosophy and evolutionary biology depicts the overall framework in which primate power structures are studied today. In this contribution, we attempt to provide a concise overview of current progress on male-female power relationships across primate societies, drawing primarily, but not exclusively, on our own recent work. We first envisage male-female power relationships in the sexual sphere, by reviewing recent work on sexual coercion in primates. We then address male-female power relationships in the social sphere, by reviewing recent work on male-female dominance relationships in primates.
8Clarifying definitions for these complex concepts that are used by scholars across several disciplines, and more traditionally in the humanities, has been an important step for studies of power in animal societies. Power is often defined as the control that one individual can exert over another individual or several other individuals. Power has been recognized as a multidimensional concept, and at least two different types of power are distinguished in animal behaviour studies: (1) resource-holding power, which reflects an individual’s ability to monopolize resources and reproductive opportunities, and (2) decision-making power, or leadership, which reflects an individual’s ability to weigh upon collective decisions (Davidian et al., 2022). Studies on decision-making power, although complex to implement in animal societies, have developed in important ways over the last two decades (Smith et al., 2016), including through studies of sex biases in leadership (Smith et al., 2020; 2021).
9In this contribution, we mostly focus on resource-holding power, which is a critical component of any individual’s competitive ability, as well as of the social structure of any social group, because it typically translates into dominance hierarchies, which rank individuals according to their relative power. Resource-holding power can be acquired via different mechanisms (Davidian et al., 2022). An obvious mechanism is coercion, which implies the use of force, or the threat of force, to rise in power. In the context of stable relationships between pairs of individuals, power based on coercion represents dominance. Dominance represents a stable, established social relationship between two individuals where one consistently wins agonistic interactions, while the other consistently loses them, which has long been assessed as a matter of routine in studies of animal behaviour (Drews, 1993). But besides coercion, individuals may also gain power by manipulating the information available to others (Bro-Jørgensen, 2011), or by leverage, in situations where they have a bargaining advantage because they possess valuable resources and services that can be exchanged (‘commodities’) but cannot be taken by force (Lewis, 2002). In the specific context of male-female relationships, females can manipulate information on their fertility to bias male attempts to control their sexuality: for example, female bonobos produce sexual swellings that are unreliable indicators of fertility (Douglas et al., 2016), preventing males from monopolizing sexual access to females who are close to ovulation. They may also use 'sex-based leverage' by trading a male’s access to sex against some resource or service (Lewis et al., 2022).
10Studies of sexual behaviour in various primates have reported on females being sexually proceptive and proactive, soliciting copulations from various males. These observations were generally interpreted as indications of female sexual freedom, whereby females triggered matings with chosen mates (Keddy-Hector, 1992; Small, 1995). However, subsequent field researchers also witnessed frequent aggression from males against females, which often appeared to be sudden, unprovoked, and not necessarily related to sexual activity, raising questions regarding the nature of male motivations to act in this way (Thompson, 2014).
11In a seminal article by Smuts and Smuts (1993), sexual coercion in animals was first introduced and defined as "the use by a male of force, or threat of force, that functions to increase the chances that a female will mate with him at a time when she is likely to be fertile, and to decrease the chances that she will mate with other males, at some cost to the female".
12It includes different behaviours such as forced copulation, when a male restrains a female physically to enforce copulation with her; harassment, when a male shows repeated aggression towards a female until she cedes and accepts mating; intimidation, when a male shows repeated aggression – over the long term and in such a way that aggression is temporally decoupled from matings – towards a female, with such aggression eventually increasing his chances of mating with her; punishment, when a male punishes a female who is mating or has mated with a rival; mate-guarding, when a male sticks to a female to prevent her from mating with any other male, and herding, when a male shows aggression towards a (fertile) female as soon as she leaves his immediate spatial proximity (Clutton-Brock and Parker, 1995).
13Males can direct aggression towards females in various contexts, which are not always visibly related to sexual activity (figure 1). One challenge is therefore to test whether such aggression represents sexual coercion. Smuts and Smuts (1993) hypothesized that male aggression towards females could be considered as sexual coercion if three predictions were supported: 1. male aggression targets fertile females; 2. male aggression is costly to the female targets; 3. male aggression increases male mating success. This hypothesis was first formally tested in wild chimpanzees (Muller et al., 2007), where all 3 predictions were found to be supported. Most male-female aggression observed in this community represented sexual intimidation, which was manifested as repeated aggression of a male towards one particular female victim, in the weeks prior to female fertility. Unlike previously known forms of sexual selection where aggression directly precedes or follows mating events, such aggression was temporally decoupled from matings, but eventually led, for the aggressor, to a better chance of mating with his victim when she became fertile. This study was important because it showed that even aggression that does not seem directly connected to sexual activity at first can still represent sexual coercion, suggesting that many forms of coercion are discreet and can easily be overlooked. Detailed studies have since replicated these findings in wild baboons (Baniel et al., 2017) and mandrills (Smit et al., 2022), confirming that sexual coercion is widespread across primate societies.
Figure 1
An adult male chacma baboon suddenly attacks an adult female for no obvious reason at Tsaobis Nature Park, Namibia. Such regular, unprovoked attacks have been shown to represent sexual intimidation, at least for the most part |
Un mâle babouin chacma adulte attaque soudainement une femelle adulte sans raison évidente à Tsaobis Nature Park, Namibie. Il a été montré que ces attaques régulières et non provoquées représentent des comportements d’intimidation sexuelle, au moins pour une part importante
Baniel et al., 2017 ; photograph/photographie : Guilhem Duvot
14The new generation of studies on sexual coercion has now moved on to better understand the ecology of coercion, and how it may vary across individuals, populations and species (Smit, 2025). Some studies show that some males are more coercive than others. For example, in chimpanzees, young males are more cooperative and less coercive than older, higher-ranking males, presumably because the former have less power than the latter (Reddy and Mitani, 2020). In mandrills, those females who maintain strong bonds with one male face more coercion than unbonded females, presumably because such male-female bonds are just as affiliative as they are coercive (Smit et al., 2023). Females that are very central in the female social network also face more coercion, probably because they are more 'exposed' to male attacks, although the reasons underlying this pattern are not yet fully clear (Smit et al., 2023). Our ongoing work is beginning to explore between-species contrasts in the occurrence of coercion by compiling literature records documenting the existence of coercive behaviour. The main difficulty is that, when sexually coercive behaviour has never been reported in a species, it is hard to ensure that it is genuinely absent, as it could instead result from an under-reporting bias. Although our analyses are not yet complete, pilot results suggest that sexual coercion is widely spread across the primate phylogenetic tree, confirming that it represents a structuring force in primate mating and social systems.
15All in all, this study shows that male-female relationships are conflictual and profoundly shaped by male coercive behaviour over female sexuality, at least in a number of well-studied species. The broader picture across the primate tree remains to be established. This view replaces previous ones describing sexually proactive females and competing males. In fact, female sexual proceptivity and mating solicitations, in the promiscuous societies of chimpanzees or chacma baboons, may reflect attempts to pacify or appease coercive males (Baniel et al., 2021), rather than being an expression of sexual freedom. This emerging picture raises questions about sexual conflict over power in the social sphere. Does the fact that males are being sexually coercive towards females in diverse primate societies reflect broader realities of primate power structures, whereby females are systematically overpowered by males?
16Historical studies on sex biases in resource-holding power in non-human primates have generally focused on biases in the outcome of male-female contests, or in other words, on sex biases in dominance relationships (Lewis, 2018). Early studies depicted male dominance over females as the default pattern across primates, with a few anecdotal cases of female-dominant species or taxa (Ralls, 1976; Smuts, 1995; Lewis, 2018), which were more or less regarded as evolutionary accidents. This picture, which included only a handful of evolutionary transitions from male to female dominance (and vice-versa), made it difficult to solve the evolutionary puzzle of female dominance using statistical approaches. Furthermore, primatologists rarely mixed males and females together in the same hierarchy on the grounds that they often compete over different resources (i.e. mates vs feeding resources) via different mechanisms (Kappeler et al., 2022). For example, male baboons fight over rank while females inherit their rank from their mother (Fischer et al., 2019). As a result, there were few intersexual dominance hierarchies, and male-female dominance relationships were sometimes established qualitatively rather than measured objectively.
17Work in the last two decades has challenged this perspective by showing that male-female dominance relationships can be meaningfully measured by the percentage of intersexual contests won by females, or alternatively, by the percentage of males dominated by an average female in a social group (Kappeler et al., 2022). Both measures correlate closely across species. With such a toolbox to hand, quantitative studies of intersexual dominance in particular species have produced a number of insights. First, they have shown that intersexual dominance can vary across different social groups of a same species (Hemelrijk et al., 2020; Surbeck et al., 2025). This suggests that male-female (or intersexual) dominance is not a binary or a fixed species trait. Secondly, they have revealed that some species had sometimes been misclassified as male-dominant, such as vervet monkeys, where females have much more power than was long thought (Young et al., 2017; Hemelrijk et al., 2020). Thirdly, these studies have shown that even within those species that represented the archetype of male dominance, such as mandrills or gorillas where adult males are two to three times heavier than females, females can occasionally win intersexual contests or dominate males, so that the power relationships are less asymmetrical than previously thought (Smit et al., 2022; Smit and Robbins, 2025). Finally, comparing measures of intersexual dominance across several species has clearly shown that the degree of female (or male) dominance varies along a continuum, with species distributed all along it and where power relationships can be fairly egalitarian between the sexes (Kappeler et al., 2022; figure 2).
18The extent of the variation in sex biases in dominance uncovered by these studies has opened up the possibility of testing hypotheses to explain the evolutionary origin of these biases, in an attempt to identify the environmental, social and life-history factors promoting female (versus male) dominance. A wide-ranging comparative study of 253 populations from 151 species has compiled quantitative data on male-female dominance (percentage of contests won by females) (Huchard et al., 2025; figure 2). In this sample, clear dominance of males, where they win more than 90% of contests over females, is observed in only 17% of the populations. The figure is almost symmetrical for clear female dominance (13%), leaving 70% of the populations with moderate (or no) sex biases in power. This study confirmed that male dominance over females does not represent the baseline across many non-human primates.
Figure 2
Distribution of dominance between sexes in primate societies. Figure adapted from Huchard et al. 2025: (A) observed percentage of confrontations won by females for each species; the black horizontal bars cover the range of values observed in different populations or studies of the same species; the colour code shows how each species was classified into a 3-level variable: Clear female dominance (winning >90% of confrontations), moderate dominance bias (each sex winning 10 to 90% of confrontations) and clear male dominance (winning >90% of confrontations). (B) Taxonomic distribution of dominance between sexes, mapped onto the primate phylogeny (credit for icons: https://www.phylopic.org) |
Répartition de la dominance entre les sexes dans les sociétés de primates. Figure adaptée de Huchard et al. 2025 : (A) pourcentage observé de confrontations remportées par les femelles pour chaque espèce ; les barres horizontales noires couvrent la gamme de valeurs observées dans différentes populations ou études de la même espèce ; le code couleur montre comment chaque espèce a été classée dans une variable à 3 niveaux : dominance féminine claire (victoire dans plus de 90 % des confrontations), biais de dominance modéré (chaque sexe remportant entre 10 et 90 % des confrontations) et dominance masculine claire (victoire dans plus de 90 % des confrontations). (B) Répartition taxonomique de la dominance entre les sexes, cartographiée sur la phylogénie des primates (crédit pour les icônes: https://www.phylopic.org)
19The study then tested five hypotheses to explain sex biases in dominance relations, and found that female-biased dominance is associated with several key factors. The first hypothesis, in the context of sexual conflict theory, posits that female social dominance over males reflects their level of reproductive control relatively to males, here mostly meaning control over when and whom to mate with. Female power is, indeed, mostly seen in species in which females can refuse matings and counter or escape male sexual monopolization attempts, such as in monogamous species, where females are of a similar size to males, or where they primarily forage in trees – all being situations in which females have more reproductive control. The second hypothesis proposes that female dominance emerges in societies where female-female competition (over energetic or reproductive resources) is particularly intense, selecting for competitive traits in females, which allow them to outcompete males. In support of this hypothesis, female dominance is observed in situations where females are as large as males and are intolerant of the proximity of other females, as in solitary or pair-living species, or in species where females defend territories. The third hypothesis proposes that lactating females may refrain from engaging in male-female conflicts and may, instead, spontaneously submit or flee, because such conflicts can be dangerous for their dependent infants. In line with this, female power was found to be more frequent in species with short lactation periods, and where mothers park offspring when they forage (as in most lemurs) instead of carrying them around with them (as in most monkeys). The fourth hypothesis proposes that female power emerges when females can form coalitions against males in order to outcompete them. In support of this hypothesis, female philopatry (or matrilocality, when females spend their lives in their natal group), which usually promotes strong female-female bonds, also promotes female dominance. However, the ability to form female-female coalitions was not found to be associated with female power, which speaks against this hypothesis and suggests that the observed association between female power and philopatry could be a consequence, rather than a cause, of male-female power relationships. In some polygynous primate societies where males are dominant over females and monopolize female groups, alpha males often force females to join their social group. Similarly, in humans, modelling work indicates that gender biases in residency appear to be absent or limited in mobile hunter-gatherers due to the lack of gender inequalities, and could have emerged during transitions to agriculture due to growing inequalities (Dyble et al., 2015). The last hypothesis posits that variation in male-female power mainly reflects variation in adult sex ratios, which affect the hierarchy via winner-loser effects (Hemelrijk et al., 2008). Under this hypothesis, female power is expected to increase as the number of males rises in a group, because more male-male contests generate injured, loser males who drop to the bottom of the hierarchy, resulting in a passive increase of the average social rank of females. Support for this hypothesis was low, because female power was not associated with a higher number of males in a group, nor with a higher proportion of male-male contests (out of all contests).
20Overall, our study found most support for the reproductive control and the female-female competition hypotheses, which are not mutually exclusive. These results further indicate that, while primate males gain power via physical force and coercion, female empowerment can rely on alternative pathways, such as reproductive strategies to gain control over matings, which affords them leverage in their power struggle with males.
21This brief overview of current knowledge on sex biases in primate power structures illustrates the rapid conceptual and empirical progress in this field. These developments have significantly transformed our understanding over the past decades. As noted in the introduction, their implications extend beyond the study of animal social behaviour, touching on broader debates about human societies.
22In response to the early primatology studies criticized by Donna Haraway (1989), a 'second wave' of more female-oriented research emerged, which included more species and was grounded in methodologies that were more quantitative and arguably more objective. This shift also involved a more diverse research community, presumably more attuned to the issues raised by Haraway, through either direct engagement with her work or with its broader influence. These more recent studies have uncovered a complex and nuanced picture of male-female power dynamics in primates, which are often more conflictual, but also more balanced, than initially assumed.
23As Haraway cautioned, drawing inferences about human societies from primate studies requires careful consideration and interdisciplinary dialogue. Nonetheless, despite her warnings against the naturalization of human behaviour, this evolving body of research suggests that tracing the evolutionary trajectory of social traits can, at times, offer the strongest empirical counterarguments to such simplistic naturalization.
24In particular, the description of female dominance as almost as common as male dominance in primate societies – and indicating that many species exhibit no clear sex bias in power – challenges traditional assumptions about the biological inevitability of gender roles. Humans do not share all the traits typical of species where males consistently dominate females. Rather, our species aligns more closely with primates displaying flexible, context-dependent hierarchies, in which individuals of either sex may attain dominance. From this perspective, framing human patriarchy as a straightforward legacy of our primate heritage appears to be not only reductive, but also misleading.
25While methods used to measure dominance differ between human and nonhuman primates (Chen Zeng et al., 2022), these results are broadly compatible with the fact that humans have spent more than 95% of their existence in hunter-gatherer societies, which are generally characterized by low or moderate levels of gender inequality (Dyble et al., 2015; Smith et al., 2021). Consequently, the prevalence of patriarchy observed in historical or contemporary human societies may represent a relatively recent feature of human social structure (Smuts, 1995; Smith et al., 2021). Interdisciplinary approaches including primatological, archeological and anthropological reports could, in combination, attempt to reconstruct the (pre)history of gender biases in power within and across past and modern human societies, as well as disentangling the relative roles played in shaping these biases by our evolutionary legacy, ecology and cultural norms and institutions.
Acknowledgements: Many thanks to the organizing team of the symposium "Hierarchy and power in human and non-human primates" of the 1850th meeting of the Société Anthropologique de Paris for their invitation and coordination, as well as to Dr Anne Le Maître and two anonymous reviewers for a smooth and constructive reviewing & editing process. Contribution ISEM no. 2026-064.