Поиск по сайту

Пишите нам: info@ethology.ru

Follow etholog on Twitter

Система Orphus

Кросс-культурный метод
Старые форумы
Не в тему

список статей

Dominance and prestige as differential predictors of aggression and testosterone levels in men

Ryan T. Johnsona, Joshua A. Burkb, Lee A. Kirkpatrickb

1. Introduction

1.1. Domain-specific self-esteem and aggression

1.2. Prestige and dominance

1.3. Dominance, prestige, and testosterone

2. Method

2.1. Participants

2.2. Measures and procedure

3. Results

3.1. Self-esteem domains and aggression

3.2. Self-esteem domains and testosterone

4. Discussion




1. Introduction

Much attention in recent years has been focused on the relationship between self-esteem (SE) and aggression, with the empirical literature evincing highly mixed results (Baumeister & Boden, 1998, Baumeister et al., 1996). Based on an extensive review of this literature, Baumeister et al. (Baumeister & Boden, 1998, Baumeister et al., 2000) conclude that SE per se is unrelated to aggression, arguing instead for a model of aggression involving ego threat and narcissism.

Alternatively, Kirkpatrick, Waugh, Valencia, & Webster (2002) have argued for an understanding of this relationship based on an evolutionary theory of SE as comprising multiple mechanisms with multiple distinct functions (Kirkpatrick & Ellis, 2001), each of which may be related to aggression in different (and adaptive) ways. Kirkpatrick etal. (2002) demonstrated empirically that domain-specific SE measures were differentially predictive of aggression in the laboratory—in some cases in opposite directions—whereas global SE was unrelated to aggression. The present research was designed to extend this line of research by (1) further distinguishing dominance vs. prestige as functionally distinct components of SE and (2) examining the relationship of these measures to aggression as well as examining the role testosterone may play in predicting SE components.

1.1. Domain-specific self-esteem and aggression

Most conventional theories in social and personality psychology conceptualize SE largely as a desired end-state—that is, something toward which people are motivated to strive. In contrast, Leary and Downs (1995) have argued that SE is a self-evaluative mechanism, analogous to the fuel gauge on a car's dashboard, that monitors the current status of one's social relations. Specifically, they argue from an evolutionary perspective that because inclusion in social groups has been crucial for survival and reproduction throughout human history, SE reflects an adaptation designed to monitor one's level of social inclusion or acceptance and to motivate corrective action when this falls below an acceptable threshold.

Kirkpatrick and Ellis (2001) extended this model by arguing that (1) because social inclusion in groups is only one of many functionally distinct adaptive problems presented by group living, SE should comprise numerous, functionally independent sociometers related to different social domains; and (2) these sociometers serve a variety of adaptive functions other than merely motivating corrective action. For example, in addition to monitoring inclusion within cooperative social relationships such as friendships and coalitions, as emphasized by Leary and Downs, people must monitor their relative standing in competitive relationships (e.g., for mates and status) in order to adaptively regulate their behavior toward others below vs. above oneself in a hierarchy (Leary & Baumeister, 2001).

This functional, evolutionary model of SE leads to the hypothesis that the relationship between SE and aggression might vary considerably across different SE domains. For example, it generally is adaptive for larger and stronger individuals to use (or threaten) aggression against weaker rivals, and for subordinates to avoid such conflicts; consequently, individuals in many species appear to assess their own and rivals' relative resource-holding power (RHP; Parker, 1974) and regulate their behavior toward competitors accordingly. Other domains of SE would be expected to relate to aggression differently: for example, high levels of social inclusion should be inversely related to aggression given the disruptive nature of aggressive behavior to cooperative social relationships.

Kirkpatrick et al. (2002) demonstrated exactly this differential pattern of empirical relationships between SE domains and aggression in college students. In their first study, a general measure of self-perceived superiority predicted higher aggression, whereas social inclusion predicted lower aggression, in a laboratory paradigm. In a second study designed to create a mate-competition context, high self-perceived mate value emerged as a positive predictor of aggression. Global SE was unrelated to aggression in both studies. Webster and Kirkpatrick (2006) have recently replicated many of these findings using both self-report and behavioral measures of aggression, with mate value and social inclusion predicting aggression differentially.

1.2. Prestige and dominance

Although Kirkpatrick et al. (2002) succeeded in distinguishing a competitive form of SE (self-perceived superiority) from a cooperative (social inclusion) form, the former construct may also be too broad. According to Henrich and Gil-White (2001), hierarchies of human status or rank (i.e., differential access to valued resources) can be based upon either of two distinct strategies or processes: Dominance refers to the attainment of status or rank through the use or threat of force, whereas prestige refers to status that is bestowed freely by others upon those displaying valued skills, knowledge, or abilities. Kirkpatrick and Ellis (2006) have since revised their model to include separate sociometers related to these distinct competitive processes.

This dominance–prestige distinction has important implications for understanding the SE–aggression link. Aggression, or the threat thereof, is inherent in dominance competition; as noted previously, high self-perceived dominance should increase the likelihood of aggression. In contrast, aggression is generally anathema to the maintenance of prestige (as with social inclusion): The display of aggression by high-prestige individuals runs the risk of driving away followers (or clientele; Henrich & Gil-White, 2001). The measure of “self-perceived superiority” employed by Kirkpatrick et al. (2002) conflates these two competitive forms of SE. The present study was therefore designed to assess self-perceived dominance and prestige separately and examine their respective predictive relationships with aggression.

1.3. Dominance, prestige, and testosterone

A secondary goal of this research was to examine the distinction between dominance and prestige using a physiological measure that may reflect one of the proximal mechanisms underlying any SE–aggression relationships. Testosterone (abbreviated T) is a steroid hormone produced and secreted in the thecal cells of the ovaries, the Leydig cells of the testes, and in cells in a portion of the adrenals (Lindzey & Korach, 1997). T is a member of a larger group of steroid molecules known as androgens, which have powerful organizational effects in multiple brain regions. Androgens are known to modulate the size and structure of brain nuclei related to emotional and cognitive processing including the medial amygdala (Johansen et al., 2004, Morris et al., 2005) and the sexually dimorphic nucleus of the preoptic area in the anterior hypothalamus (Morris, Jordan & Breedlove, 2005). Generally speaking, androgens are thought to act as signals to induce cell death or proliferation in these regions; however, the exact signaling pathway is unclear.

Empirical relationships between T and aggressive behavior are conflicting but well documented (Book et al., 2001, Harris et al., 1996, Olweus et al., 1988; see also Archer, Graham-Kevan, & Davies, 2005; see Giammanco, Tabacchi, Giammanco, Di Majo, & Guardia, 2005 for a review). Where direct measures of aggressiveness are available (i.e., nonhuman animal subjects), testosterone is related to aggressive behavior (Lumia, Thorner, & McGinnis, 1994). However, for obvious ethical reasons, research investigating testosterone and aggression in human subjects is more restricted and perhaps as a result the relationship between the two variables is not as certain (O'Connor, Archer, Hair, & Wu, 2002; see also Giammanco et al., 2005). Additionally, testosterone has been shown to fluctuate in response to environmental cues such as competition or challenge (Neave & Wolfson, 2003, Salvador et al., 2003, Wagner et al., 2002; see Archer, 2006 for a review). Furthermore, in both human (Newman, Seller, & Josephs, 2005) and nonhuman subjects (Briganti et al., 2003, Mitchell & Wilson, 1987), testosterone appears to interact or respond to social rank. Considering the relevance of behaviors such as competition, social rank, and aggression to dominance, we expected T levels to be differentially related to measures of dominance and prestige with individuals high in dominance exhibiting higher testosterone levels than individuals high in prestige.

2. Method

2.1. Participants

Participants were 72 men and 67 women (mean age=19.1) enrolled in introductory psychology courses at the College of William and Mary. Of these, 43 men (mean age=19.2) volunteered for the saliva collection portion of the study. (Women were excluded from the testosterone portion of the study due to the extreme complexity of female hormonal fluctuations.) Participants were required to indicate, on a list provided, any medications or nutritional supplements (known to influence T-assay results) that they had been taking recently; however, no participants had to be removed from the sample on this basis. All participants were thoroughly debriefed at the end of the study and received course credit for their participation.

2.2. Measures and procedure

Each participant completed the following paper-and-pencil measures: the Global Self-Esteem Scale (Rosenberg, 1965); the Aggression Questionnaire (AQ; Buss & Perry, 1992); and the Self-Perceived Social Status Scale (SSSS; Buttermore, James, & Kirkpatrick, 2005), with all responses given on a Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). The SSSS yields separate scores for self-perceived dominance and prestige. Examples of items scoring for dominance include “I demand respect from members of my peer group” and “Others believe they can push me around” (reverse-scored). Examples of items scoring for prestige include “Others recognize me for my contributions to my social group” and “Others do not value my opinion” (reverse-scored). The AQ produces scores for physical aggression, verbal aggression, anger, and hostility. Each participant also completed a short form containing demographic information. Questionnaire order was randomized within each packet, and participants were given approximately 30 min to complete the materials. Because Tlevels are cyclical, all testing was conducted between noon and 4:00 p.m. (Table 1).

Table 1.

Descriptive statistics

Scale Range Min Max Mean S.D.
Men (n=71)
Hostility aggression 37 9 46 27.66 8.31
Verbal aggression 27 8 35 20.97 5.46
Anger aggression 31 13 44 24.82 7.16
Physical aggression 38 14 52 30.41 9.34
Dominance 5 1.5 6.5 3.79 0.91
Prestige 4.1 2.7 6.8 5.11 0.77
Global SE 2 3.4 5.4 4.26 0.40
Women (n=67)
Hostility aggression 35 10 45 28.38 7.71
Verbal aggression 27 7 34 18.45 5.72
Anger aggression 28 14 42 24.13 6.43
Physical aggression 34 13 47 24.85 7.40
Dominance 4.6 1.5 6.1 3.72 0.97
Prestige 5.3 1.3 6.6 5.01 0.80
Global SE 1.7 3.4 5.1 4.23 0.35

For male participants in the second phase, saliva was collected using a passive drool method. Salimetrics LLC (State College, PA, USA) performed the assays using an enzyme immunoassay. Analysis was performed in duplicate with a sensitivity range of 1.5 to 360 pg/ml. Inter-assay correlation was 0.989. A mean T score for each sample was produced by averaging the two assays, and this value was used for statistical analysis.

3. Results

3.1. Self-esteem domains and aggression

Correlations were computed between the various aggression and SE measures, and each of the four aggression scores served as the dependent variable in a separate multiple regression equation with dominance, prestige, and global SE as predictors. Data for men and women were analyzed separately. The results for each equation, along with corresponding zero-order correlations, are shown in Table 2 for men and Table 3 for women.

Table 2.

Multiple regression analysis and correlations predicting aggression from SE domains in men (n=71)

B SE B β r
Dominance 3.51 1.02 0.38** 0.33**
Prestige −3.86 1.16 −0.36** −0.28*
Global SE 1.54 2.25 0.07 0.14
Verbal aggression
Dominance 3.14 0.66 0.53** 0.47**
Prestige 0.51 0.82 0.07 −0.18
Global SE 0.38 1.83 0.02 −0.01
Dominance 4.57 0.83 0.58** 0.53**
Prestige −0.44 0.95 −0.05 0.05
Global SE −3.38 1.85 −0.19 −0.06
Physical aggression
Dominance 4.33 1.19 0.42** 0.36**
Prestige −0.73 1.36 −0.06 0.01
Global SE −4.68 2.64 −0.20 −0.11

*p<.05; **p<.01.

Table 3.

Multiple regression analysis and correlations predicting aggression from SE domains in women (n=66)

B SE B β r
Dominance 3.51 1.02 0.38** 0.35**
Prestige −3.86 1.16 −0.36** −0.21
Global SE 1.54 2.25 0.07 0.30*
Verbal aggression
Dominance 3.14 0.66 0.53** 0.56**
Prestige 0.51 0.82 0.07 0.26*
Global SE 0.38 1.83 0.02 0.18
Anger aggression
Dominance 3.72 0.72 0.56** 0.52**
Prestige −2.10 0.90 −0.26* 0.00
Global SE 4.03 2.00 0.22 0.27*
Physical aggression
Dominance 4.03 0.86 0.53** 0.54**
Prestige −0.76 1.07 −0.08 0.15
Global SE 3.13 2.39 0.15 0.25*

*p<.05; **p<.01.

For men, hostility was positively correlated with dominance, negatively related to prestige, and unrelated to global SE. When the three SE measures were assessed simultaneously in a multiple regression equation, dominance remained a significant positive predictor, prestige remained a significant inverse predictor, and global SE remained unrelated. The other three aggression subscales—verbal aggression, physical aggression, and anger—were all positively related to dominance both in zero-order correlations and in multiple regressions. Prestige and global SE were unrelated to verbal aggression, physical aggression, and anger.

For women, hostility and anger were positively correlated with dominance and global SE. However, in a multiple regression equation, global SE was not a significant predictor of hostility or anger, while prestige emerged as a significant inverse predictor. Physical aggression was correlated with dominance and global SE. However, the relationship with global SE was no longer significant when analyzed using multiple regression. Prestige was unrelated to physical aggression. Verbal aggression was positively correlated with both prestige and dominance, but only dominance remained a significant predictor of verbal regression in a multiple regression equation.

3.2. Self-esteem domains and testosterone

T levels (mean=297.9 pg/ml, S.E.=17.9 pg/ml) were unrelated to time of testing (r=−.20, p=.20) or age (r=−.03, p=.88), as expected given the restricted age range and controlled sample collection time.

Results for correlational and multiple-regression analyses predicting T levels from dominance, prestige, and global SE are presented in Table 4. As predicted, T levels were differentially predictive of dominance and prestige. Prestige emerged as a significant inverse predictor of T, whereas dominance was unrelated to T. Instead, it was global SE that emerged as a positive predictor of T.

Table 4.

Multiple regression analysis and correlations predicting testosterone levels from SE (n=43)

B SE B β r
Dominance −15.97 19.65 −0.12 −.06
Prestige −46.62 22.62 −0.31** −.23
Global SE 123.94 47.96 0.39** .30


In a final set of analyses, we used multiple regression to predict T levels from the four aggression subscales. None of the scales was significantly correlated with aggression nor related to it in a multiple regression equation (all p's>.05).

4. Discussion

In partial support of our predictions based on Henrich and Gil-White's (2001) theoretical distinction between dominance and prestige, in conjunction with the Kirkpatrick & Ellis, 2001, Kirkpatrick & Ellis, 2006 model of domain-specific SE, our results showed prestige and dominance to relate differentially to both self-reported aggression in men and women and to T levels in men. In both sexes, dominance was positively related, while prestige was either inversely related or unrelated to the various self-report measures of aggression. In men, prestige was also inversely related, but dominance unrelated, to T levels. These results extend the previous findings of Kirkpatrick et al. (2002) and demonstrate the importance of distinguishing these two functionally distinct forms of self-evaluation with respect to aggression.

Our prediction that prestige would be inversely related to aggression was confirmed for only two of the four subscales of the Aggression Questionnaire, hostility and anger. Note that these scales do not measure aggression per se, but rather personality traits that might ordinarily predispose people toward aggression. Prestige was not related to the two behavioral (verbal aggression and behavioral aggression) subscales of the AQ. It stands to reason that a person with high prestige would be expected to be less (dispositionally) angry and hostile than others; someone who is admired, respected, and valued presumably has little to be angry or hostile about. Why, though, does high prestige not lead either to lower levels of behavioral aggression?

An important part of the answer to this question may involve the nature of our measures of physical aggression, which are self-reports of the general tendency to engage in aggressive behavior, presumably as averaged across time and context by participants when responding to the questions. Such trait-level measures might well produce different results than a situation-specific behavioral measure of aggression in response to provocation, as in previous studies by Kirkpatrick et al. (2002). The frequency with which someone behaves aggressively over time is a function of numerous factors beyond their probability of doing so in a particular situation, such as the frequency with which he or she faces such provocations. Thus, one plausible explanation for the pattern of results observed here is that high-prestige people are indeed less likely to aggress in response to any given provocation, but they aggress as often (on average, across time) as low-prestige persons because they are provoked more often. In other words, high-prestige persons may be aggressing no more than low-prestige persons despite the fact that they are more frequently provoked. Given that prestige represents a form of high status, it seems entirely possible that this might be the case: Although they may have their own loyal clienteles, prestigious persons are in constant competition with other prestigious persons for followers and as such might be expected to regularly face challenges for their high-status positions from competitors—in much the same way that those at the top of a dominance hierarchy do.

This interpretation might also explain why prestige emerged as a significant inverse predictor of T levels. If, as Henrich and Gil-White (2001) propose, prestige processes reflect a very recent evolutionary development, the psychological mechanisms designed to implement them might operate by suppressing aggression (among high-prestige individuals) that otherwise would be produced by the more primitive dominance system.1 To illustrate with an example to which journal readers should easily relate, consider what happens when a high-profile speaker at an academic conference, or a professor in a classroom, is criticized by an audience member. Physically assaulting the questioner would be a sure-fire way to undermine the speaker's (prestige-based) status and end his or her career. Fortunately, this kind of response rarely occurs: such individuals invariably suppress momentary urges to lash out physically against rivals.

Consistent with this line of reasoning, Suarez, Kuhn, Schanerg, Williams, and Zimmerman (1998) found that men characterized by low hostility demonstrated a smaller increase in testosterone and were less likely to report feelings of anger and negative affect in response to harassment than men high in hostility. Combining these observations with our own results, we suggest that reduction of T levels may reflect the activity of an aggression–suppression mechanism in high-prestige individuals, designed to override the aggression that would otherwise be produced in high-status individuals by the more primitive dominance system. This hypothesis is admittedly post hoc, but suggests a promising avenue for future research.

The question remains, however, why we did not find the predicted relationship between dominance and T levels. Previous findings have been mixed regarding the relationship between T levels and aggression, and the relationship is still under considerable debate in the literature. For example, a recent meta-analysis found only a weak positive relationship between T and aggression (Archer et al., 2005). Consistent with our own previous suggestion, some researchers have argued that the relationship is not detectable with trait-level questionnaire measures, but instead is context dependent and observed only during periods of challenge in circumstances relevant to reproduction (Archer et al., 2005).

Researchers have also begun to examine other physiological factors that may play a role in regulating aggression. For example, low levels of serotonin metabolites are associated with increased aggression, especially impulsive aggression (Bjork et al., 1999, Lee & Coccaro, 2001). Additionally, male mice treated with nonaromatizable T and subsequently given a serotonin receptor agonist showed reduced aggressive behavior. However, the effect was dependent both upon the type of agonist used and the dose administered (Cologer-Clifford, 1997). These findings suggest a complex serotonin×T interaction in controlling aggressive behavior, with serotonin acting as a buffer that attenuates aggressive behavior. Such an interaction might explain many of the conflicting findings from research examining T and aggression alone.

One final finding that bears discussion concerns global SE, which we found to be positively related to T levels in men. One possible explanation stems from research identifying a positive role for androgens in cognitive functions and mood (Alexanderson & Christiansen, 2004), and studies examining the behaviors and attitudes of androgen-replacement therapy patients have found particularly interesting results. Androgen supplementation, in addition to increasing libido and muscle mass, is associated with an improvement in general sense of well-being in women (Goldstat, Briganti, Tran, Wolfe, & Davis, 2003). In hypogonadal men, androgen supplementation results in a reduction on self-report measures of tension, anger, and fatigue (O'Connor et al., 2002). However, O'Connor et al. (2002) are careful to point out that in the case of androgen replacement, increases in self-esteem may be closely related to restoration of sexual function. Androgen treatment in the opposite direction seems to produce the expected effect. Men receiving androgen deprivation therapy to treat prostate cancer are eight times as likely to be diagnosed with clinical depression. This increase was not related to worsening of disease, type of androgen deprivation, medical response to androgen deprivation, or chemotherapy (Pirl, Siegel, Goode, & Smith, 2002). However, like the link between testosterone and aggression, this relationship is complex and not clearly defined. Additional studies have found no relationship between androgen treatment and mood (Gray et al. 2005). Nevertheless, it remains plausible that T may have a relationship with higher-order functions, such as mood and well-being, which may serve to influence aspects of self-esteem.

In summary, our results demonstrate that self-appraisals of dominance and prestige are differentially related to both self-reported aggression and physiological processes that might regulate aggressive behavior. Given the multi-faceted and highly complex nature of all of the variables involved—aggression, self-esteem, and physiology—much research will be needed to untangle the various empirical relationships and explicate the functional processes involved. Furthermore, it should be noted that these findings represent a narrow age range and modest hormone sampling size (n=43). Additionally, testosterone was collected only from men for logistical reasons. Thus our interpretation is limited in its scope as testosterone may play a very different role in modulating female aggression and competitive hierarchy behavior.

Finally, one specific implication of our findings from this research is that understanding processes that inhibit aggression, as in the context of prestige competition, may prove to be as important as understanding those processes that cause or facilitate aggression. Moreover, it seems clear that such research will need to be highly interdisciplinary in nature, bringing together methods and measures from social psychology and neuroendocrinology within an evolutionary theoretical framework.


This research and preparation of this manuscript were facilitated by the College of William and Mary through a Minor Research Grant to Ryan T. Johnson, a faculty research assignment to Lee A. Kirkpatrick, and internal funds to Joshua A. Burk. We are grateful to Jenée James and Nicole Buttermore for help in data collection and other assistance. We also thank two anonymous reviewers of a previous version of this manuscript for their unusually detailed and insightful comments.


Alexanderson & Christiansen, 2004 1.Alexanderson P, Christiansen C. The aging male: Testosterone deficiency and testosterone replacement. An update. Atherosclerosis,. 2004;173:157–169. Abstract | Full Text | Full-Text PDF (203 KB) | MEDLINE | CrossRef

Archer, 2006 2.Archer J. Testosterone and human aggression: An evaluation of the challenge hypothesis. Neuroscience and Biobehavioral Reviews,. 2006;30:319–345. MEDLINE | CrossRef

Archer et al., 2005 3.Archer J, Graham-Kevan N, Davies M. Testosterone and aggression: A reanalysis of Book, Starzyk, and Quinsey's (2001) study. Aggression and Violent Behavior,. 2005;10:241–261.

Baumeister & Boden, 1998 4.Baumeister RF, Boden JM. Aggression and the self: High self-esteem, low self-control, and ego threat. In:  R. Green,  E. Donnerstein editor. Human aggression: Theories, research, and implications for social policy (pp. 111–37). San Diego, CA: Academic Press; 1998;.

Baumeister et al., 2000 5.Baumeister RF, Bushman BJ, Campbell WK. Self-esteem, narcissism, and aggression: Does violence result from low self-esteem or from threatened egotism?. Current Directions in Psychological Science,. 2000;91:26–29.

Baumeister et al., 1996 6.Baumeister RF, Smart L, Boden JM. Relation of the threatened egotism to violence and aggression: The dark side of self-esteem. Psychological Review,. 1996;103:5–33. CrossRef

Bjork et al., 1999 7.Bjork JM, Dougherty DM, Moeller FG, Cherek DR, Swann AC. The effects of tryptophan depletion and loading on laboratory aggression in men: Time course and a food constricted control. Psychopharmacology,. 1999;142:24–30. MEDLINE | CrossRef

Book et al., 2001 8.Book AS, Starzyk KB, Quinsey VL. The relationship between testosterone and aggression: A meta-analysis. Aggression and Violent Behavior,. 2001;66:579–599.

Briganti et al., 2003 9.Briganti F, Seta DD, Fontani G, Lodi L, Lupo C. Behavioral effects of testosterone in relation to social rank in the male rabbit. Aggressive Behavior,. 2003;293:269–278.

Buss & Perry, 1992 10.Buss AH, Perry M. The Aggression Questionnaire. Journal of Personality and Social Psychology,. 1992;633:452–459.

Buttermore et al., 2005 11.Buttermore N, James J, Kirkpatrick L. Prestige and dominance: Toward a self-report measure of two distinct pathways to status. In: Poster presented at the Evolutionary Psychology Preconference, Society of Personality and Social Psychology, Austin, TX, 2005. 2005;.

Cologer-Clifford, 1997 12.Cologer-Clifford A. Steroidal modulation of 5-ht(1a) and 5-ht(1b) agonist effects on offensive aggression in male mice. Dissertation Abstracts International. B. The Sciences and Engineering,. 1997;57:6784.

de Waal, 1982 13.de Waal FBM. Chimpanzee politics: Power and sex among apes. London: Jonathon Cape; 1982;.

de Waal & Luttrell, 1989 14.de Waal FBM, Luttrell LM. Toward a comparative socioecology of the genus macaca: Different dominance styles in rhesus and stumptail monkeys. American Journal of Primatology,. 1989;19:83–109. CrossRef

Giammanco et al., 2005 15.Giammanco M, Tabacchi G, Giammanco S, Di Majo D, La Guardia M. . Testosterone and aggressiveness. Medical Science Monitor,. 2005;11:RA136–RA145. MEDLINE

Goldstat et al., 2003 16.Goldstat R, Briganti E, Tran J, Wolfe R, Davis SR. Transdermal testosterone therapy improves well-being, mood, and sexual function in premenopausal women. Menopause,. 2003;105:390–398.

Gray et al., 2005 17.Gray PB, Singh AB, Woodhouse LJ, Storer TW, Casaburi R, Dzekov J, et al. Dose-dependent effects of testosterone on sexual function, mood, and visuospatial cognition in older men. Journal of Clinical Endocrinology & Metabolism,. 2005;90:3838–3846. CrossRef

Harris et al., 1996 18.Harris JA, Rushton JP, Hampson E, Jackson DN. Salivary testosterone, self-report aggressive and pro-social personality characteristics in men and women. Aggressive Behavior,. 1996;225:321–331.

Henrich & Gil-White, 2001 19.Henrich J, Gil-White FJ. The evolution of prestige: Freely conferred deference as a mechanism for enhancing the benefits of cultural transmission. Evolution and Human Behavior,. 2001;22:165–196.

Johansen et al., 2004 20.Johansen JA, Jordan CL, Breedlove SM. Steroid hormone masculinization of neural structure in rats: A tale of two nuclei. Physiology and Behavior,. 2004;83:271–277. MEDLINE | CrossRef

Kirkpatrick & Ellis, 2001 21.Kirkpatrick LA, Ellis BJ. Evolutionary perspectives on self-evaluation and self-esteem. In:  G. Fletcher,  M. Clark editor. The Blackwell handbook of social psychology: Vol. 2: Interpersonal processes (pp. 411–36). Oxford, England: Blackwell; 2001;.

Kirkpatrick & Ellis, 2006 22.Kirkpatrick LA, Ellis BJ. What is the evolutionary significance of self-esteem?. In:  M. H. Kernis editors. Self-esteem issues and answers: A source book of current perspectives (pp. 334–9). New York: Psychology Press; 2006;.

Kirkpatrick et al., 2002 23.Kirkpatrick LA, Waugh CE, Valencia A, Webster G. The functional domain-specificity of self-esteem and the differential prediction of aggression. Journal of Personality and Social Psychology,. 2002;825:756–767.

Leary & Baumeister, 2001 24.Leary MR, Baumeister RF. The nature and function of self-esteem: Sociometer theory. In:  M. Zanna editors. Advances in Experimental Social Psychology (pp. 1–62). San Diego, CA: Academic Press; 2001;.

Leary & Downs, 1995 25.Leary MR, Downs DL. Interpersonal functions of the self-esteem motive: The self-esteem system as a sociometer. In:  M. H. Kernis editors. Efficacy, agency, and self-esteem (pp. 123–44). New York: Plenum; 1995;.

Lee & Coccaro, 2001 26.Lee R, Coccaro E. The neuropsychopharmacology of criminality and aggression. Canadian Journal of Psychiatry,. 2001;46:35–44.

Lindzey & Korach, 1997 27.Lindzey J, Korach K. Comparative endocrinology. In:  P. M. Conn,  S. Melmed editor. Endocrinology: Basic and clinical principles (pp.47–62). Totowa, NJ: Humana Press Inc; 1997;.

Lumia et al., 1994 28.Lumia AR, Thorner KM, McGinnis MY. Effects of chronically high doses of the anabolic androgenic steroid, testosterone, on inter-male aggression and sexual behavior in male rats. Physiological Behavior,. 1994;55:331–335.

Mitchell & Wilson, 1987 29.Mitchell JA, Wilson MC. The behavioral and toxicological effects of testosterone pellet implantation. Physiological Behavior,. 1987;41:427–432.

Morris et al., 2005 30.Morris JA, Jordan CL, Breedlove SM. Sexual differentiation of the vertebrate nervous system. Nature Neuroscience,. 2005;7:1034–1039. MEDLINE | CrossRef

Morris et al., 2005 31.Morris JA, Jordan CL, Dugger BN, Breedlove SM. Partial demasculinization of several brain regions in adult male (XY) rats with a dysfunctional androgen receptor gene. Journal of Comparative Neurology,. 2005;487:217–226. MEDLINE | CrossRef

Neave & Wolfson, 2003 32.Neave N, Wolfson S. Testosterone, territoriality, and the ‘home advantage’. Physiology & Behavior,. 2003;782:269–275.

Newman et al., 2005 33.Newman ML, Sellers JG, Josephs RA. Testosterone, cognition, and social status. Hormones and Behavior,. 2005;47:205–211. CrossRef

O'Connor et al., 2002 34.O'Connor DB, Archer J, Hair WM, Wu FCW. Exogenous testosterone, aggression, and mood in eugonadal and hypogonadal men. Physiology and Behavior,. 2002;75:557–566. MEDLINE | CrossRef

Olweus et al., 1988 35.Olweus D, Mattsson A, Scfshalling D, Low H. Circulating testosterone levels and aggression in adolescent males: A causal analysis. Psychosomatic Medicine,. 1988;50:261–272. MEDLINE

Parker, 1974 36.Parker GA. Assessment strategy and the evolution of fighting behavior. Journal of Theoretical Biology,. 1974;47:223–243. MEDLINE | CrossRef

Pirl et al., 2002 37.Pirl WF, Siegel GI, Goode MJ, Smith MR. Depression in men receiving androgen deprivation therapy for prostate cancer: A pilot study. Psycho-Oncology,. 2002;11:518–523. MEDLINE | CrossRef

Rosenberg, 1965 38.Rosenberg M. Society and the adolescent self-image. Princeton, NJ: Princeton University Press; 1965;.

Salvador et al., 2003 39.Salvador A, Suay F, Gonzalez-Bono E, Serrano MA. Anticipatory cortisol, testosterone and psychological responses to judo competition in young men. Psychoneuroendocrinology,. 2003;283:364–375.

Stammbach, 1988 40.Stammbach E. Group responses to specially skilled individuals in a Macacca fascicularis group. Behavior,. 1988;107:241–266.

Suarez et al., 1998 41.Suarez EC, Kuhn CM, Schanberg SM, Williams RB, Zimmermann EA. Neuroendocrine, cardiovascular, and emotional responses of hostile men: The role of interpersonal challenge. Psychosomatic Medicine,. 1998;60:78–88. MEDLINE

Wagner et al., 2002 42.Wagner JD, Flinn MV, England BG. Hormonal response to competition among male coalitions. Evolution & Human Behavior,. 2002;236:437–442.

Webster & Kirkpatrick, 2006 43.Webster GD, Kirkpatrick LA. Behavioral and self-reported aggression as a function of domain-specific self-esteem. Aggressive Behavior,. 2006;32:17–27.

a Neuroscience Program, Michigan State University, East Lansing, MI 48911, USA

b Department of Psychology, College of William and Mary, Williamsburg, VA 23187, USA

Corresponding author. Neuroscience Program, Michigan State University, Giltner Hall Office 208, East Lansing, MI 48911, USA.

1 Contrary to Henrich & Gil-White's (2001) claim that prestige processes are uniquely human—being rooted in unique human abilities for direct “infocopying”—nonhuman primates have been reported to have different “styles” of status (de Waal, 1982, de Waal & Luttrell, 1989, Stammbach, 1988), and deference to individuals with unique skills does exist in other species (Stammbach, 1988).

PII: S1090-5138(07)00040-2