Masculinity and Aggression in South Africa

By Russell Luyt

Introduction
Mainstream ('malestream') academia has by and large neglected the critical examination of men. This is certainly true in the study of violence. Foster notes that men have been disproportionately exposed to its ill effects either as perpetrators or victims [1]. South African history paints a vivid portrait of the close association between men and violence.

Fortunately interest in the study of masculinity has blossomed over the last two-and-a-half decades. Feminist theoretical advances and the politics of the gay liberation movement have encouraged the development of Men's Studies. The current study identified with this development. In doing so it attempted to problematise the relationship between masculinities and violence. Gender theorising that stresses the importance of power proved useful - in this case Connell's [2] often-cited concept of hegemonic masculinity. This concept:

"provides a way of explaining that though a number of masculinities coexist, a particular version of masculinity holds sway, bestowing power and privilege on men who espouse it and claim it as their own" [3:608].

Hegemonic masculinity serves to sustain male power both in relation to women and subordinate masculinities. All men therefore benefit within this gender order. As such most support it despite their frequent departure from the hegemonic ideal. That is to say every male incorporates elements that are both consistent and contradictory with hegemonic masculinity. In particular differences emerge along social axes such as 'race', class, and sexuality. These differences act as the basis for contestation surrounding its definition [2]. Hegemony is therefore never total. Literature suggests that aggression and/or violence often accompanies such contestation, where the enactment of either, serves as a marker of a man's masculinity[4].

This research sought to explore the variable endorsement of hegemonic masculinity by South African men and its relationship to aggression. It adopted a constructionist perspective. Male aggression is not considered essential. It finds perpetuation in society, in part, due to its close association with the social construct of masculinity. Literature also suggests that age [5] and education [6] are related to aggression. Accordingly this research explored the core hypothesis that: low age and education, together with strong hegemonic masculine approval, are predictive of high aggression whilst high age and education, together with weak hegemonic masculine endorsement are predictive of low aggression. Eight variables were initially included in the analysis: age, education, masculine toughness score, masculine control score, masculine sexuality score, total masculinity score, transformed frequency aggression score, and transformed duration aggression score.

Method
Stratified purposive sampling across age and education provided a means with which to isolate 432 suitable participants. Involvement was restricted to Afrikaans, English and Xhosa speaking male South Africans. Participant age ranged between 15 and 87 years old and averaged approximately 37. The Male Attitude Norms Inventory-II [MANI-II] [7] was utilised as a multi-dimensional measure of masculinity ideology. Its three sub-scales served to guide construction of a multi-dimensional model of masculinity. This model provided a blueprint for the construction of three similar sub-scales. Principal factor analytic procedures (Communalities Multiple R²) rendered single-factor solutions in each case. Each also demonstrated firm overall [Total Scale (a = 0.90)] and individual [Toughness Sub-Scale (a = 0.69); Control Sub-Scale (a = 0.86); Sexuality Sub-Scale (a = 0.74)] internal reliability. An adaptation of the Buss Aggression procedure [8] was used to assess individual predisposition toward aggression. Buss's original experimental procedure guided the construction of an 'aggression machine'. Individuals were required to participate in a task in which they supposedly administered electric shocks to a research confederate. An individual's willingness to administer 'shock' treatment served as a measure of their aggression. 

Results
Multiple correlation indicated significant relationships (a = 0.05) between most of the variables and by-and-large confirmed theoretical assumptions. However the relationship between age and the four masculine variables was unexpected. That is to say age was significantly positively correlated to masculine toughness (r = 0.19), control (r = 0.31), sexuality (r = 0.34), and the total scale (r = 0.34). This suggested that as men got older they increasingly supported traditional masculinity.

Three predictor variables were found to contribute meaningfully to a regression model accounting for aggression using the transformed frequency aggression score. Both the masculine sexuality score (Beta = 0.12; p < 0.05) and education (Beta = - 0.12; p < 0.05) made a significant contribution to the model. Age (Beta = - 0.10; p = 0.08) failed to do so despite making a useful contribution to the model as a whole. The overall regression model accounted for 3.8 % of the variance in aggression [F(3.36) = 4.62; p < 0.001]. Only main effects were considered in this analysis.

Two independent samples t-tests were conducted. The first explored the hypothesis that men diverging in age, education, and their support for traditional masculinity (total masculinity score) would differ significantly in aggressive response. Comparison between mean aggression scores revealed significant results (t (56) = 2.40; p < 0.01). Young, poorly educated men who strongly supported traditional masculinity (M = 139.74; Valid N = 23) were found to aggress more readily than older, well educated men who endorsed traditional masculinity to a lesser extent (M = 110.17; Valid N = 35). The second t-test was undertaken in order to explore the surprising finding that age was significantly positively correlated to the four masculine variables. Comparison between mean total masculinity scores also produced significant results (t (253) = -6.49; p < 0.01). Young Xhosa men (M = 107.39; Valid N = 128) were found to support hegemonic masculinity to a significantly lesser extent than older Xhosa men (M = 119.20; Valid N = 127). Post hoc comparisons were not conducted. An analysis of variance may prove interesting at a later stage.

Discussion
The research hypothesis was supported. Low education and high endorsement of traditional masculine sexuality was seen to be significantly predictive of aggression. Low age, although not uniquely contributory to aggression, was found to contribute meaningfully to the overall predictive model. This model only accounted for a small proportion of the variance in aggression. Klineberg usefully reminds us "it is impossible to find a single cause of all forms of violence. We are dealing here clearly with a multidimensional phenomenon, and our understanding of it demands that we keep many facets simultaneously in mind" [9:122]. Future studies should explore the contribution of additional variables in predicting aggression. Nevertheless the current findings are easily understood. Some studies suggest that aggression and/or violence plays a pivotal role in the lives of young disempowered males [10]. With few alternatives available, young males adopt these behaviours, and in so doing reinforce their status as 'true' men.

Further avenues for research emerged. The discovery that 'sexuality' surfaced as a significant predictor of aggression rather than masculine 'toughness' or 'control' holds particular interest. The traditional masculine norm of 'sexuality' arguably involves processes of social categorisation to a greater extent than either 'toughness' or 'control'. It is plausible that individual male violence may be encouraged through such processes. The relationship between social categorisation and the masculine norm of 'sexuality', in encouraging aggression, should be explored further. Social dominance theory [11] may be a worthwhile conceptual tool in an undertaking of this kind.

The correlation result, indicating that as men got older they increasingly supported traditional masculinity, also deserves further exploration. It could be argued that the highly traditionalist culture among older 'black'-African men [12], who comprised a substantial proportion of the entire sample (65.7%), may have contributed toward this finding. Even so a future study comparing traditional masculine endorsement across culture and age may prove insightful.

Acknowledgements
This paper draws on research undertaken under the auspices of the Department of Psychology at the University of Cape Town. Research funds were partially made available through the University. Particular thanks go to Jenny Luyt for her unreserved encouragement, Gillian Finchilescu for her statistical advice, and Don Foster for his overall supervisory guidance and support.

References
1. Foster, D. (1997). Perpetrators of Gross Violations of Human Rights. Cape Town, South Africa: Report made to the Truth and Reconciliation Commission.

2. Connell, R.W. (1987). Gender and Power: Society, the Person and Sexual Politics. Stanford, California: Stanford University Press.

3. Morrell, R. (1998). Of boys and men. Journal of Southern African Studies, 24, 605-630.

4. Breckenridge, K. (1998). The allure of violence: Men, race and masculinity on the South African goldmines, 1900-1950. Journal of Southern African Studies, 24, 669-693.

5. Turner, B.F. (1992). Gender differences in old age in ratings of aggression/assertiveness. Current Psychology: Research and Reviews, 11, 122-127.

6. Harris, M.B. & Knight-Bohnhoff, K. (1996). Gender and aggression: II. Personal aggressiveness. Sex Roles, 35, 27-42.

7. Luyt, R. (2005). The Male Attitude Norms Inventory-II: A measure of masculinity ideology in South Africa. Men and Masculinities, forthcoming

8. Buss, A.H. (1961). The Psychology of Aggression. New York: John Wiley and Sons.

9. Klineberg, O. (1981). The causes of violence. In J. Domenach, H. Laborit, A. Joxe, J. Galtung, D. Senghaas, O. Klineberg, J.D. Halloran, V.P. Shupilov, K. Poklewski-Koziell, R. Khan, P. Spitz, P. Mertens, E. Boulding (Eds.), Violence and its Causes (pp. 111-123). Paris: Unesco.

10. Mooney, K. (1998). 'Ducktails, flick-knives and pugnacity': Subcultural and hegemonic masculinities in South Africa, 1948-1960. Journal of Southern African Studies, 24, 753-774.

11. Sidanius, J. & Pratto, F. (1999). Social Dominance: An Intergroup Theory of Social Hierarchy and Oppression. Cambridge: Cambridge University Press.

12. Carton, B. (2001). Locusts fall from the sky: Manhood and migrancy in KwaZulu. In R. Morrell (Ed.), Changing Men in Southern Africa (pp. 129-140). Pietermaritzburg, South Africa: University of Natal Press. 

* Russell Luyt is with the Institute of Social Psychology at the London School of Economics, London, UK.


Download PDF version [115Kb] Back to main page