Individual differences in the study of human aggression

Excessive concern for simplicity has often led us to underestimate the complexity of aggression. Over the years, several scales related to aggression have been identified and validated both on adults and on children. Ultimately, two main higher order dimensions, emotional responsivity and proneness to aggression, repeatedly emerged in a series of second-order factor analysis. These dimensions are rather similar to two of the big five factors of personality: neuroticism/emotinal stability and hostility/agreeableness, respectively. Several studies corroborate the plausibility of a higher level bidimensional conceptualization of aggression that, while preserving the heuristic validity of looking for more specific constructs of aggression, can link the field of aggression to the field of personality. © 1994 Wiley-Liss, Inc.     

Behavioral changes over several successful agonistic encounters between male mice: Effects of type of “standard opponent”

This study assessed whether two types of non-aggressive “standard opponents” (“intact” and “anosmic” group-housed males) produced similar behavioral changes in isolated OF1 male mice given several experiences of victory. Experimental groups confronted either intact or anosmic opponents every two days until they had completed four encounters. The behavioral changes were recorded using a detailed ethologically inspired analysis. These changes were clearly different depending on the opponent type. When intact opponents were used, experimental subjects increased the time spent in digging, non-social exploration, explore from a distance, and attack over encounters, but showed decreased time spent in threat and a decreased latency to the first attack. In encounters with anosmic opponents, only declines in the latencies to threat and attack were noted. Moreover, the experimental groups differed in their behaviors over encounters. Those confronting intact opponents spent less time in social investigation, more time in explore from a distance and threat, and showed a shorter latencies to threat and attack than counterparts confronting anosmics. These results suggest that, although both types of “standard opponents” are similar in their non-aggressiveness, they elicit rather different behavioral responses in their adversaries. These findings provide additional support for the view that the type of opponent used in studies on intermale aggression is of paramount importance. Indeed, the use of different types of “standardized non-aggressive opponents” appears to be an important source of variability between studies. © 1994 Wiley-Liss, Inc.     

Variation in aggressive behavior and anatomo-physiological correlates generated by crowding without physical contact in the house mouse

Behavioral, physiological (i.e., endocrine), and anatomical consequences of crowding in mice were studied in a situation where animals were in auditory, visual, olfactory, and tactile contact but restrained from full physical interactions, to prevent overt aggression. Males that cohabited with females undisturbed by neighboring conspecifics showed greater propensity to attack same-sex intruders and had higher plasma testosterone levels than did their “crowded” counterparts, that is, males cohabiting with females and housed adjacent to other male/female pairs. In this respect, the latter animals resembled submissive males. However, a significant increase in weight of androgen-dependent target organs (i.e., seminal vesicles and preputial glands) was found in crowded males. These data indicate that despite the observed inhibition of social aggression these males are not physiologically comparable (homologous) to male mice that experienced defeat and the stress of submission during fighting. The intriguing possibility that different conversion pathway of testosterone are accelerated, as a result of social communication, in males living in these two environments and the behavioral implications of these possibilities are discussed. Finally, the parental behavior of crowded animals, although not freely interacting with each other, was disrupted, causing a marked decrease in reproductive success. In this situation a high incidence of infanticide of their offspring by both parents was observed, whereas this behavior was virtually absent in non-crowded male/female pairs.     

Relative deprivation: A mechanism for the ostracism–aggression link

In five studies, we tested whether ostracism triggers feelings of relative deprivation and whether relative deprivation accounts for the impact of ostracism on aggression. Relative to participants who recalled either inclusive or neutral experiences, participants who recalled ostracism experiences reported higher levels of relative deprivation (Study 1). Furthermore, the feeling of relative deprivation mediated the effect of ostracism on aggression (Studies 2, 3a, and 3b). Framing ostracism as an experience of nondeprivation weakened the connection between ostracism and aggression (Study 4), which suggests an effective way of reducing aggression following an ostracism experience. Together, these findings highlight the significance of relative deprivation in shaping people's responses toward ostracism.     

An empirically derived scoring system for intermale aggression in mice

Scoring systems used to assess intermale aggression have been characterized by arbitrary scales and wide variability in the behaviors selected for measurement. The use of such different systems severely limits the ability of investigators to make meaningful comparisons among studies and indicates that there is a need for a common, statistically derived evaluative system for intermale aggression. We measured the frequency and duration of five major components of agonistic behavior exhibited by intact males toward olfactory bulbectomized stimulus males and then analyzed the data using a number of univariate and multivariate procedures. The results were used to generate two statistically based scoring systems, one a short-form index and the other a composite index for more detailed studies of aggression. It is hoped that these statistically derived systems will be adopted by other investigators to increase methodological congruence in the field.     

Effects of midbrain central gray lesions on spontaneous and electrically induced aggression in the rat

Large electrolytic lesions were placed in the midbrain central gray of male rats. Their effects on hypothalamically induced aggression, switch-off behaviour, and locomotion were investigated. A number of these animals were also tested for territorial intermale aggression in order to compare electrically induced and spontaneous aggression. Large lesions resulted in an increase of the current threshold to induce aggression by hypo-thalamic stimulation. Smaller, but still quite large, lesions decreased the threshold current for hypothalamic aggression. After the operation a decrease in the threshold for switch-off was present, both in the experimental and the control group. Current thresholds for locomotion were decreased after the lesions only in the experimental group. Spontaneous aggression was temporarily decreased after the lesion. No indication was found that other behavioural elements of the animal were distorted by the lesion. The parallel between the effects on spontaneous and electrically induced aggression makes it attractive to ascribe a role to the neural circuit of hypothalamus and central gray in territorial aggression. However, even with large lesions the animals were still capable of fighting, hence the central gray is not indispensable. An attempt was made to explain the differential effects that differently sized central gray lesions have on hypothalamic aggression.     

Separate neural sites for d-amphetamine suppression of mouse killing and feeding behavior in rats

The present study was conducted to investigate several possible neural sites for d-amphetamine's effect on mouse killing and feeding behaviors. d-Amphetamine (10, 20, and 30 μg) injected into each lateral ventricle, suppressed mouse kiling, food, and water intake in a dose-dependent manner. Bilateral adminstration of d-amphetamine (20 μg) into the central amygdaloid nucleus abolished mouse killing behavior but did not affect feeding and drinking. By contrast, bilateral amphetamine injections into the substantia nigra, or into the ventral region of the caudate nucleus, did not suppress mouse killing behavior, but significantly decreased food and water intake. The lateral hypothalamus was sensitive to d-amphetamine injections, which suppressed mouse killing and food intake as well as water intake. d-Amphetamine injections into the nucleus accumbens produced inconsistent effects on mouse killing and feeding. Our observations suggest a differentiation of the neural sites that mediate feeding from those underlying mouse killing behavior.     

Visual aspects of centrally elicited attack behavior in the cat: “Patterned reflexes” associated with selection of and approach to a rat

It has been previously suggested that the electrical brain stimulation which elicits quiet-biting attack in the cat actively affects the way the central nervous system processes visual and tactile information concerned with the reflexes involved in the terminal aspects of attack. In order to examine the effects of brain stimulation on a nonterminal aspect of attack – the stimulated cat's selection of and approach to a rat – cats were implanted with attack-eliciting electrodes in both the lateral hypothalamus and the midbrain ventral tegmental area. These cats were then tested in an 8-ft-long cage, one end of which was divided into three, 2-ft-long parallel compartments, whose openings faced the end of the cage from which the cat commenced its approach. An anesthetized rat was placed at the back of one compartment, a bowl of food at the back of another compartment, and the third compartment contained no object. It was found in the first experiment that the attack elicited by nearly all electrodes was selectively directed at the rat. However, the success of the cat in finding the compartment containing the rat varied dramatically for different electrodes in the same cat. Further, these differences were stable and did not change as the cat gained experience with the task. The results suggested that the stimulation of different brain sites in the same cat differentially affected the visual neural mechanisms involved in guiding a cat to a rat. Previous studies have also suggested that the effects of brain stimulation which elicits quiet-biting attack are largely lateralized to the side of the brain stimulated. In order to determine if the effects of stimulation on the neural mechanisms mediating the visually guided approach of a cat to a rat were also lateralized, attempts were made in a second experiment to disrupt the visual input to one side of the brain by unilaterally transecting the optic tract. It was found that this manipulation interfered with the visually guided selective approach to a rat, if the cat was stimulated through hypothalamic or mid-brain electrodes ipsilateral to the optic tract transection, but not if the hypothalamic or midbrain stimulation was on the contralateral (visually intact) side of the brain. However, any final interpretation of the results was confounded by the finding in all of these cats of a complex syndrome of neglect of all contralateral sensory information.     

Naturalistic versus experimental approaches to aggression; theoretical and methodological issues

Various strategies to study human aggression are described, accompanied by a discussion of four major issues. It is concluded that the three major goals of research, precise measurement, realism, and generality, cannot be achieved in a single study or series of studies using only one paradigm. Investigators of a particular problem need to develop a research program that incorporates a number of strategies; hypotheses generated “in the street” need to be tested in the laboratory and vice versa.