The Crowd Around the Fight:
An Analysis of the Acceptance of Violence
Joe Chacon and Elise Vigil, The University of New Mexico
Sociology 481: Research Methods, Art St. George

Taking a look at reasons why people would be acceptant of violence, we examine trends found among persons who approve of intra-sexual male violence. First, sex differences are examined concerning biological and social explanations of such attitudes. Hypotheses are made according to past research in which males are the aggressors who more likely favor intra-sexual competition due to status and sexual fitness. We used data from the General Social Survey to find relationships between attitudes favorable towards male-male violence and other factors representative of social and sexual status. A regression analysis shows low correlation of violent attitude with sexual status and a significant relationship with social status.

Given the societal concern regarding violence, we have decided to take an approach through the social, psychological, and biological aspects of human understanding of the phenomenon. With differential rates of males as aggressors and females as victims, as well as the racial disparity in incarceration statistics, it is important that such a social problem as violence be better understood along the lines of sex, sexuality, and social factors determining social status.

In regard to violence as it is observed by males and females apart, it has been found that regardless of punishment history, attitudes towards violence are inherently learned. These include violent and aggressive behaviors through interpersonal relationships and the modeling process (MacIntyre et al., 1995). Along the same lines, Cultural Spillover Theory states that violence and aggression that is legitimized will engender violence and aggression in many areas of life. Our hypothesis was to differentiate between male versus female attitudes in regard to violent behaviors with males more likely to favor male-male violence. To validate this hypothesis, research by MacIntyre et al. was assessed, and it was found that the results obtained in a study of 145 women and 95 men showed that a significant gender difference existed in the attitudes towards violence and aggression. To elaborate, it was also found that while the mean violence score for men was higher than that of women (men, 2.18 and women, 1.90) there was no significant difference in scores between sex groups (Smith et al., 2001 p.45). Smith et al. went on to theorize that it is not gender differences that determine the favorability of violence and aggression, but that more people regardless of their interpersonal style are moving towards a view that does not condone violence in any reasonable sense.

Despite gender differences in the favorability of male-male violence, it is often thought in society that violence is highly correlated with young, black inner-city males as opposed to white males. It is believed that black males are more prone to violence due to the arrest numbers and incarceration rates of this subgroup in the United States. However, it can easily be stated that the disproportion could be due to mere discrimination by the criminal justice system (Stark, 1993). The opposing view shows that through self-report surveys of crime victims it is reported that if you are a victim of violent crime, your assailant is almost three times more likely to be white than black. This contradicts the belief that the typical assailant is a black male. It is also seen that according to the National Crime Survey, the rate of black males being victims of aggravated assault is about equal to the rate for white males (32 per 1,000 and 31 per 1,000, respectively).

To quickly discuss the in’s-and-out’s of evolutionary theory is to discuss more than a hundred year’s work. However, simply put and in association with male aggression, natural selection asserts that male aggression is a protective adaptation, and thus a survival solution; sexual selection considers aggression to be a solution to mating pressures. Sexual selection works to promote certain qualities in one sex, as chosen by the opposite, which then leads to the thriving of such genes related to the attractive qualities (Buss 2003, p. 3). Aggression can be considered a highly correlated characteristic of dominance, which it turns out, is often cited as an attractive male quality (Johnston, Hagel, Franklin, Fink, & Grammer 2001; Mazur, Halpern, Udry 1994). According to such evolutionary theories, through the pressures of sexual selection, male aggression has coevolved with selected male dominance.

To assert that male aggression is not only prevalent but part of the male psychology, Adam Fox found that men tend to assess their fighting ability more often than their female peers do (1997). Although more frequent assessors of ability, men fail to be as accurate as women (1997). The findings illustrate a predisposition to aggressive thought. Likewise, men tend to fantasize about homicide much more than women do (Kenrick & Sheets 1993). Men are more detailed in their homicidal fantasies than women and tend to imagine strangers as their victims more often than women, who consider lovers and family members as victims in their much shorter fantasies (p. 235, 237-238). This helps to confound the significance of UCR data showing that strangers have a 24% chance of victimization in relation to an 8% chance of victimization by a family member, with the exception of wives (Crime in the United States 2003). Male aggression is most often directed towards other men, especially unknown men, who it can be argued, tend to be status competitors in some degree (Wilson & Daly 1985; Hilton, Harris, & Rice 2000). If not an acquaintance or stranger, then victims tend to be friends or sexual partners (Crime in the United States 2003). Thus, aggression is a product of natural and sexual selection.

Heterosexual females report higher attraction rates for masculine faces on men more so than for feminine faces on men (Johnston et al. 2001; Mazur et al. 1994). However, this changes over the course of the menstrual cycle, as shown by Johnston et al. When at high risk for pregnancy, females prefer the more masculine face. The attractive male, in turn, denotes a healthy gene carrier to the female due to testosterone markers of broad chin, pronounced brow ridges and pronounced cheekbones. It is important to note that women at the low risk end of menstruation report higher attraction rates to the feminized male faces than to the overly masculine faces. Johnston et al. suggest that the negative attributes of testosterone markers become “associated with dominance, unfriendliness, and…traits [indicating] threat, volatility, manipulation, coerciveness, and selfishness” (p. 262). But, as suggested by Mazur et al’s title, Dominant Looking Male Teenagers Copulate Earlier, dominance is selected for by the female sex (1994). While attractiveness was not totally correlated with dominance, dominance was still a better predictor of sexual activity, as well as an elevated social status within the high school setting (p. 92-93). However, as was noted by Mazur et al., “dominant men… enjoy… greater sexual access” than non-dominant males. Therefore, there is an obvious correlation between dominant features, social status, and sexual status that should further be investigated.

While substantial evidence shows a male proclivity towards aggressive assessments, thoughts, and behavior, such a proclivity is a double standard in human evolution. Male teenagers report that they more often aggress against other male teenagers than they do against female teenagers (Hilton et al. 2000). Hilton et al’s study also shows that males will report male-male aggression but underreport male-female aggression (p. 9-10). Socially, it almost seems acceptable for males to be aggressive, especially when the aggressed are other males. However, when females are the victims, even the perpetrators tend to conceal their aggressive tendencies to researchers (Hilton et al. 2000). With the social expectations of male aggression, it should be considered that such male activity tends to be found attractive and correlative with physical looks. As Hilton et al. illustrates, male psychology reflects this with male teenagers inferring less insult and injury in male-female aggression than females report considering such aggression.

Male aggression is a product of many social factors, including environment, biological predisposition such as disorder, and even levels of testosterone (Mazur et al. 1994, p. 93). The social significance of male aggression is articulated by the differential rates of homicide, indicating a male-dominated social dilemma. With the understanding of the evolutionary adaptedness of male aggression, for the sake of survival and reproduction, aggression in males needs further study in regards to it’s status-assuming capabilities and it’s correlation with physical attractiveness. However, it shouldn’t be understood in terms of a female desire for negative male aggression. Male aggression often indicates risk-taking behaviors and a status driven character. Rather, as Buss makes clear with his discussions of social status, size, and strength, dominance remains attractive for it’s inferences of health, status, and ability to provide resources for mother and child (p. 25-27, 38-40). Not as attractive or selected for it’s indications of depravity and violence, dominance was most likely selected for it’s ability to circumvent other men who tried to dominate women sexually and aggressively (Buss 2003, p. 40). Thus, dominance and aggression coevolved with the evolution of aggression being a protective quality of dominance and the resource securing behaviors producing attraction in females.

Based on the literature reviewed, favorable versus non-favorable views of violence can be based on general demographical data for a given population. This is relevant given the amount of work establishing findings of several different reasons for the function and expression of male violence. Using GSS data, we would like to further substantiate such research with the ample amount of variables made available. Thus, attitudes favorable towards violence will be examined along the lines of several variables, such as race, gender, and number of sex partners.

We used data from the General Social Survey, and utilizing the variables we chose to analyze, the longitudinal data is representative of respondents from 1988 to 1994. Over the course, respondents were asked questions ranging from general demographics to personal perspectives on various social phenomena. We were concerned with ten variables descriptive of demographics and their relationship with a specific attitude.

Using SPSS, we recoded many variables (as noted below) and applied all pertinent variables in a hierarchical multiple regression model. In addition, we provide two cross tabulations of respondent’s sex and the number of male and female sex partners for a better understanding of the respondents and possible relationships with sexuality and sexual fitness.

Listed below are the dependent and independent variables as noted by their unique variable code names, recoded variable names if applicable, and are then immediately followed by the survey question asked of respondents.

It should be noted that while HITOK as a variable has the ability to lend to specific questions following a positive response, our preliminary analyses found that all answers to questions such as respondents approving of a male hitting a stranger or a robber or a man that has just hit a woman or child were highly correlated. Thus, HITOK is a safe and representative variable of the various situations in which it can be found that male-male violence is appropriate. The use of HITOK will work to describe male-male violence in general terms, regardless of specificity.

Using our nine possible relationships of variables, we hypothesize the following given previous findings in the areas of male aggression and attitudes favorable towards violence in order to substantiate contemporary social, biological, and psychological explanations of male violence:

Our subject pool represents respondents surveyed between the years of 1988 and 1994. This accounts for our relatively small group size of 2, 291 respondents. Most of our respondents approved of a man punching or hitting another adult man; had an education level of at least 11 years up to about two years of post-secondary education; were unemployed at the moment; were white, married, and female; ranged in the age groups of 33 to 57 years of age; had about 1 through 5 male and/or female sex partners; and had no more than one child at the most.

Concerning our correlation output, only three of our seven independent variables significantly correlated with our dependent variable, HITOK. “Dummy for working” correlated with HITOK (r = .120). “Education levels” correlated with HITOK, although negatively it remains to hold a significant relationship at the p = .01 level (r = -.150).

The race of a respondent correlated with HITOK, significantly (r = .116). Being male least correlated with HITOK (r = .002). Thus, employed people approved less or declined response when asked if they approved of a man hitting another man. Those respondents with less education than the mean population were more inclined to approve of male-male violence, and whites were more inclined to approve of male-male violence as well. However, the insignificance of male as an indicator shows that females were more likely to approve of such violence.

Model One shows that 2.2% of the variability is explained by the variable of education level. Model Two shows an increase in the variability to 2.9% with the addition of “Dummy for Working” to the model. Model Three increases once again to 3.9% variability explained with the addition of race as a variable. Model Four increases to an overall 4.4% variability explained by education, employment, race, and finally, age.

As explained in the analysis of correlation, working, education, and race show significant relationships with HITOK. The addition of age to the model did not change the significance of education level from Model Three to Four. In fact, respondent’s race increased with significance from Model Three to Four, from a beta of .106 to a beta of .115. All variables in Model Four are significant at the .01 level.

Hierarchical multiple regression substantiated our hypotheses about the relationships between education level, age, race, and employment status. Thus, in concordance with the literature, race has a significant relationship with violent attitudes. Non-whites are just as likely or less likely than whites to be victims or aggressors (Stark 1993). We also found that a person’s education level is a correlational factor in determining favorability towards violence. In our data set, those with the higher levels of education were less likely to approve of male-male violence, as our hypothesis predicted. Likewise, working individuals favored male-male violence less than those who were unemployed at the time of data collection.

However, only three of our nine hypotheses were substantiated by the analysis of our data set. Given the literature, our hypotheses should have held true. Although, our analyses found there was no significant relationships between sex, number of sex partners, marital status, and number of children and approving of male-male violence. Sex and the approval of violence was not significant along the lines of being male; the number of sex partners did not determine approval and the number of offspring did not create a significant relationship.

Given our findings, it is essential to further study the relationships between sex, gender, and reproduction and the approval of a man hitting another man. For now, social factors such as those determining social status, like education and employment, maintain their significance in predicting negative affect of male-male violence.