The usual explanation for the link between testosterone levels and aggression is that testosterone interacts with androgen or oestrogen receptors. During the critical time period, testosterone sensitises certain neural circuits in the brain. This allows for the effects that testerone have which last into adulthood. Testosterone can affect the effect of neurotransmitters, e.g. how fast they move and the amount of neurotransmitter released. It appears to act upon serotonergic synapses, lowering the amount of serotonin available for synaptic transmission. The significance of this is that serotonin inhibits aggression, so less serotonin means more aggression. To prove this, Simpson (2001) carried out an experiment on rhesus monkeys where he gave them serotonin reuptake inhibitors, leading to a marked decrease in aggression. This has been implicated on humans also and has the same effect (Simpson, 2001).
Sapolsky has argued that although there is clearly some kind of link between levels of testosterone and aggression, this tells us nothing about individual differences in how people are affected by testosterone, why men are exceptionally more aggressive than women and does not explain why sometimes the highest levels of aggression are found in those with not an exceptionally large amount of testosterone. The relationship between testosterone and aggression is not straightforward. Does testosterone cause aggression? Does aggression increase testosterone secretion? Does neither have an effect on the other (i.e. a third variable is involved)?
Bernhardt et al (1998) has shown that merely watching participants win or lose in sports competitions increases testosterone secretion in those who are interested in the game. Indeed, Bernhardt did an experiment where he measured participants’ testosterone levels before and after sports events. Men who watched their team lose had a decrease of testosterone levels by 20% and those whose team won showed a 20% increase in testosterone levels. Clearly, testosterone does not always cause behaviour/emotion, but rather the opposite can occur too. This somewhat dents the argument that hormones explicitly cause aggression.
Klinesmith et al (2006) showed how testosterone levels change as a result of behaving aggressively. Male participants held either a gun or a children’s toy as part of an experiment. Those who held the gun behaved three times as aggressively as those who handled the toy. Thus, it would seem that aggressiveness can be affected via testosterone by environmental stimuli.
In conclusion, it seems that the effect of hormones such as testosterone are still very contrived. It would be reductionist to assume that hormones indefinitely cause aggression, but there does seem to be an effect. Any conclusions on causation must be reviewed with caution.
No effort was made in these studies to distinguish the different types of aggression, i.e. hostile, pro-social and instrumental aggression. This may be important in determining where the causes of aggression lie. Testerone appears to be implicated in only certain forms of aggression e.g. inter-male aggression, but plays no part in predatory aggression (Simpson, 2001).
There is more than just testosterone involved in implicating aggression. This is seen in non-human animals, e.g., many animals display different levels of aggression depending on the time of year and breeding season, with a corresponding increase in testosterone levels.
There are also methodological problems with testosterone-aggression research. It is not particularly easy to measure testosterone levels in animals. The castration subtraction and replacement technique is also not perfect. It is unethical and also castration subtraction and replacement techniques affect many hormone systems, so identifying the effects of one hormone, namely testosterone, is effectively impossible. Animal studies are not necessarily generalisable to humans – Van de Poll and van Goozen (1992) write that ‘there is a danger of triviality or even misleading simplification in many of our extrapolations and animal models.”
One psychological explanation of aggression is Social Learning Theory, proposed by Bandura. Social Learning Theory or ‘SLT’ proposes that we can be influenced by environmental factors. It is often called observational learning, because we learn the behaviours of others – as well as learning through our own experience -- by observing them and incorporate these behaviours into our own behaviour patterns.
Bandura proposed that we all are born without aggression but we model the behaviours, attitudes and emotional reactions of others. The theory suggests that aggression is a learned behaviour, like so many other behaviours. SLT proposes that behaviour that is reinforced (i.e. rewarded) will be repeated and learned; aggression that is associated with a reward is likely to be incorporated into our behaviour. Learning can also occur indirectly, through observing other people, i.e. by vicarious experience.
As Bandura himself says, “Learning would be exceedingly laborious, not to mention hazardous if people had to rely soley on the effects of their own actions to inform them what to do. Fortunately most human behaviour is learned observationally through modelling: from observing others one forms an idea of how new behaviours are performed, and on later occasions this coded information serves as a guide for action.”
There are four components to SLT – four processes that result in people incorporating behaviour that they have witnessed/experienced into their own typical behaviour.
Attention: someone can learn through observation as long as they are actually observing the person by paying attention to them.
Retention: One must incorporate the memory of what they have observed (i.e. the behaviour) into their long term memory.
Production: an individual must be capable of reproducing the model’s behaviour so they must possess the physical capability of the modelled behaviour. So, if the behaviour was kicking, you must be able to generate enough energy to move your legs.
Motivation: if the person feels that if they carry out the behaviour there will be positive reinforcements (e.g. some kind of praise or gift) then they are much more likely to carry it out. The reinforcement must be in the form of an outcome that they value; the model being similar to the observer and being a powered and admired role model is usually required for the incorporation of the behaviour; the model must be seen as similar to the learner, e.g. the same sex/age/interests; the task must be neither too difficult nor too easy to imitate; the observer must have confidence in their own abilities. Bandura also believed that behaviour reinforced by family members has the greatest effect on the learner.
Bandura, Ross and Ross carried out a study to observe the effects of social or vicarious learning. The experiment involved showing children models carrying out aggressive behaviour on a Bobo Doll. The results were that children who had witnessed an aggressive model punching the doll scored significantly higher on levels of aggression than those in the control group, where their model did not punch the doll at all. Thus it would seem that aggression is influenced by us modelling our behaviours on others.
Of course, the study can be criticised for demand characteristics, as the situation was staged and the children may have felt that they were supposed to hit the doll, both because that is what it was designed for and if they were in the aggressive condition, because they felt that they were supposed to copy the behaviour. Other criticisms include the fact that only western children were used, no adults or people from other cultures, so the study suffers from age- and culture-bias.
Overall, SLT has a good reputation amongst psychologists.
It has useful applications for psychologists dealing with controlling people’s aggression and other behaviours.
It has been supported by numerous studies, including the Bobo doll study. Other studies of Bandura’s have found that watching cartoon aggression produces as much aggression in children as viewing live or filmed aggressive behaviour.
It makes cognitive sense that environmental experiences would influence our social learning of violence, especially as children.
Some have criticised SLT for being reductionist however, as it fails to account for individuals’ biological factors and the differences of indivuals due to genetic, neurological and learning differences (Jeffery, 1990).