Specification: Neural and hormonal mechanisms in aggression, including the roles of the limbic system, serotonin and testosterone. 

In psychology, aggression is behaviour intended to cause harm to another person who is motivated to avoid that harm. Psychologists distinguish between two types. Reactive (hostile) aggression is angry, impulsive retaliation in the heat of the moment, such as punching someone who has just shoved you. Proactive (instrumental) aggression is cold, planned and unprovoked, and is used to achieve a goal, such as a planned robbery. The biological mechanisms on this page explain reactive aggression far better than proactive aggression.

Biological explanations argue that aggression is produced by the structure and chemistry of the brain and by hormones. This page explains two neural mechanisms (the limbic system, and serotonin acting on the orbitofrontal cortex) and two hormonal mechanisms (testosterone and progesterone). As you read, notice that these are not separate explanations. They work together as one system: the limbic system generates the aggressive impulse, while serotonin and hormones affect how well that impulse is controlled.

Key terms

• Limbic system: a group of subcortical brain structures, including the amygdala, hypothalamus and hippocampus, involved in emotion, motivation, memory and the assessment of threat.

• Amygdala: an almond-shaped structure in the temporal lobe that attaches emotional significance to sensory information and is activated by threatening stimuli.

• Orbitofrontal cortex (OFC): part of the prefrontal cortex, just behind the eyes, involved in impulse control, anticipating consequences and regulating social behaviour. It normally restrains the amygdala.

• Serotonin: a mainly inhibitory neurotransmitter with a calming effect, involved in mood and impulse control.

• 5-HIAA: the substance produced when serotonin is broken down. Low levels in cerebrospinal fluid indicate low serotonin activity.

• Testosterone: an androgen (male sex hormone) produced mainly by the testes, linked to dominance and aggression.

• Progesterone: a sex hormone produced mainly by the ovaries, whose levels rise and fall across the menstrual cycle.

• Cortisol: a stress hormone released by the adrenal glands.

The limbic system

The limbic model proposes that aggression, especially reactive aggression, is controlled by a set of structures deep in the brain. Each structure has a different role, and aggression results from the way they work together.

• The amygdala rapidly judges whether incoming sensory information is a threat, before we are consciously aware of it. An over-reactive amygdala tags harmless cues, such as an accidental bump, as threatening and produces a stronger fight-or-flight response. As a result, the person is more likely to respond aggressively.

• The hypothalamus regulates the autonomic nervous system. When the amygdala signals a threat, the hypothalamus activates the sympathetic nervous system, raising heart rate and blood pressure and preparing the body to fight.

• The hippocampus compares the current situation with memories of similar past events. This allows a person to recognise that a situation is not really dangerous. If the hippocampus is impaired, the amygdala responds without this context, so harmless situations may be treated as threats.

• The orbitofrontal cortex (OFC), part of the prefrontal cortex, normally inhibits the amygdala. It acts as a "brake" on emotional impulses.

The neural explanation can therefore be summarised as a balance between two systems. The amygdala is a fast emotional system that generates the aggressive impulse. The OFC is a slower control system that decides whether to act on it. Aggression is most likely when the amygdala is highly reactive and the OFC exerts weak control over it.

Key study: Gospic et al. (2011)

Aim: To investigate the role of the amygdala in reactive aggression.

Procedure: Participants played a lab-based bargaining game (the ultimatum game) while their brains were scanned using fMRI. During the game they received unfair offers designed to provoke them, and could react aggressively by rejecting them. Before the game, some participants took a benzodiazepine, a drug that reduces arousal of the autonomic nervous system, while others took a placebo.

Findings: Aggressive reactions to unfair offers were associated with a fast and heightened response in the amygdala. Participants who had taken the benzodiazepine made about half as many aggressive reactions, and their amygdala activity was reduced.

Conclusion: The amygdala plays an important role in reactive aggression: when its activity is reduced, aggressive behaviour is reduced too.

The orbitofrontal cortex and serotonin

Serotonin is a neurotransmitter: a chemical released by one neuron that crosses the synapse and affects whether the next neuron fires. It is mainly inhibitory, so it makes neurons less likely to fire. In the brain's emotional system, normal levels of serotonin in the OFC help it to restrain the amygdala. In other words, serotonin is part of the OFC "brake", and it has a calming effect that helps a person control emotional, impulsive reactions.

The low serotonin hypothesis states that low levels of serotonin reduce self-control. Denson et al. (2012) explain that decreased serotonin disrupts the normal neural activity of the OFC. As a result, the person has less self-control and behaves more impulsively and aggressively. They also find it harder to anticipate the consequences of their actions. This means that a minor provocation, which most people would ignore, is more likely to produce an aggressive reaction. Because this explanation is about a loss of self-control, it fits reactive aggression much better than planned, proactive aggression.

Supporting research comes from offenders. Virkkunen et al. (1994) measured 5-HIAA, the substance produced when serotonin is broken down, in the cerebrospinal fluid of violent offenders. They found significantly lower levels of 5-HIAA in impulsive violent offenders than in violent offenders whose violence was not impulsive. This suggests that low serotonin is linked specifically to impulsive violence, which is exactly what the hypothesis predicts.

Testosterone

Across human and animal studies, males are generally more physically aggressive than females, which suggests that male hormones play a role. Testosterone is an androgen produced mainly by the testes, and in small amounts by the ovaries and adrenal glands. It rises sharply in adolescence, peaks in early adulthood and then declines with age, which matches the age pattern of male violence. Testosterone is thought to increase aggression by acting on the brain areas described above: it may make the amygdala more reactive and reduce the regulating influence of the OFC. This means that testosterone strengthens the impulse and weakens the brake at the same time.

Animal research supports this. Studies reviewed by Giammanco et al. (2005) show that castration, which removes the main source of testosterone, reduces aggression in animals, while testosterone injections restore it. Research on humans also supports the link. Dolan et al. (2001) found a positive correlation between testosterone levels and aggressive behaviour in a sample of 60 male offenders in UK maximum-security hospitals, who had personality disorders.

The relationship is more complex than "more testosterone, more aggression". Mazur's (1985) biosocial model of status proposes that testosterone changes in response to social situations, especially competitions for status, and that these changes then influence later aggressive behaviour. Testosterone and the social environment therefore affect each other.

Progesterone

Progesterone is a sex hormone produced mainly by the ovaries. Its level rises and falls across the menstrual cycle. Progesterone is thought to have a calming effect, so lower levels may be linked to greater irritability and aggression. This suggests that hormones are relevant to female aggression as well as male aggression.

Ziomkiewicz et al. (2012) measured progesterone in daily saliva samples from 122 women across the menstrual cycle and asked them to rate their own premenstrual symptoms, including aggression and irritability. They found a negative correlation between progesterone and self-reported aggression: the lower a woman's progesterone, the more aggression she reported. This supports the view that progesterone may help to regulate aggression in women. However, because the study is correlational and relies on self-report, it cannot show that low progesterone causes aggression.

Evaluation

Research supports the role of the limbic system, but extrapolating from animals to humans is difficult. Kluver and Bucy (1939) removed parts of the temporal lobes, including the amygdala, in rhesus monkeys. The monkeys became unusually tame, with greatly reduced fear and anger responses. In humans, Wong et al. (1997) used PET scans of 31 male offenders with schizophrenia in a maximum-security hospital and found reduced activity in the front of the temporal lobes (the region that contains the amygdala) compared with non-offender controls. Together these studies suggest that the temporal lobe and amygdala region is involved in aggression. However, the human prefrontal cortex is far larger than a monkey's and exerts much more control over the limbic system, so the monkey findings may overstate the limbic system's role in people. Wong's all-male, secure-hospital sample also limits how far the findings generalise. On balance, the evidence suggests the limbic system is involved in aggression, but it cannot tell us how important it is for aggression in the general human population.

Aggression cannot be explained by the limbic system alone. The amygdala's output is regulated by the orbitofrontal cortex. Coccaro et al. (2007) found that OFC activity is reduced in people with psychiatric disorders that feature aggression. Because the OFC plays a crucial role in impulse control and the inhibition of aggressive behaviour, reduced OFC activity removes this restraint, and these individuals behave more aggressively. This suggests that a complete neural explanation must include the OFC as well as the limbic system. The explanation is therefore strongest when it is framed as a balance between the amygdala and the OFC, rather than as a limbic system explanation on its own.

Drug studies suggest a causal role for serotonin, but the effects may differ between the sexes. Mann et al. (1990) gave 35 healthy adults dexfenfluramine, a drug that acts on the serotonin system, and found a link between serotonin and hostility and aggression in the male participants but not the female participants. Berman et al. (2009) gave participants either a placebo or paroxetine, a drug that increases serotonin activity, before a lab game in which they gave and received electric shocks in response to provocation. The paroxetine group gave fewer and less intense shocks, and this effect was clearest in participants with a history of aggression. Because serotonin was deliberately manipulated, Berman et al.'s study provides stronger evidence of cause and effect than correlational studies such as Virkkunen et al. However, Mann's sex difference suggests the mechanism may not be the same in men and women. Assuming that findings from male participants apply equally to women would be an example of beta bias, so the serotonin explanation may be more valid for male aggression than for female aggression.

Aggression may change serotonin, rather than the other way round. Ferrari et al. (2003) allowed rats to fight at the same time each day for ten days. On the eleventh day, when no fight took place, the rats' serotonin levels had fallen and their dopamine levels had risen in anticipation. This suggests that the experience of aggression can itself change brain chemistry. Low serotonin may therefore be a consequence of aggression as well as a cause, which challenges the assumption that biology is always the starting point. This does not necessarily disprove the serotonin explanation. Instead, it suggests a two-way relationship in which low serotonin makes aggression more likely and aggression then lowers serotonin further, which supports an interactionist view of nature and nurture.

Animal research shows that testosterone can cause aggression, but it may not generalise to humans. Wagner et al. (1979) castrated mice and found that their aggression fell; when the mice were later injected with testosterone, their aggression returned. Because testosterone was manipulated, this provides experimental evidence of a causal role, which correlational human studies such as Dolan et al. cannot provide. However, human aggression is far more strongly influenced by social and cognitive factors, such as learned norms and the ability to think about consequences, than mouse aggression is. The hormonal explanation may therefore overstate the role of testosterone in human aggression, and the animal findings should be applied to people with caution.

Hormones interact with social situations and with each other. Mehta and Josephs (2006) tested Mazur's biosocial model by asking more than 60 men to compete in pairs on a number-tracing task, measuring testosterone before and after. Among the men who lost, 73% of those whose testosterone rose chose to compete against their opponent again, compared with only 22% of those whose testosterone fell. This suggests that the competition changed testosterone, and testosterone then shaped how the men responded. In addition, the dual-hormone hypothesis (Carré and Mehta, 2011) proposes that high testosterone leads to aggression only when cortisol is low, because high cortisol blocks testosterone's influence. Together, these findings suggest that testosterone on its own is a poor predictor of aggression. A model that combines testosterone, cortisol and the social situation is likely to be more valid than a simple testosterone explanation.

The explanations are deterministic and reductionist. Neural and hormonal explanations suggest that aggression is caused by brain structures, neurotransmitters and hormones, which are outside a person's control. This raises difficult questions for the justice system: if violence was caused by low serotonin or a damaged OFC, is the person fully responsible? However, most people with low serotonin or high testosterone are never violent, which suggests that these factors raise the risk of aggression rather than determining it. This is soft rather than hard determinism. The explanations are also reductionist, because they explain a complex social behaviour at the level of amygdala activity, serotonin and testosterone. This precision is useful, since it identifies mechanisms that treatments such as paroxetine can target, but it cannot explain why people with similar biology behave very differently in different cultures and situations.

Overall

There is strong and varied evidence that the limbic system, serotonin and hormones such as testosterone are involved in aggression, especially reactive aggression. Studies that manipulate biology, such as Gospic et al. (2011), Berman et al. (2009) and Wagner et al. (1979), show that changing these factors changes aggressive behaviour. However, much of the human evidence is correlational and based on male samples, and research such as Ferrari et al. (2003) and Mehta and Josephs (2006) shows that experience can change biology. The most balanced conclusion is that neural and hormonal mechanisms create a predisposition to reactive aggression, which interacts with the social situation, rather than being its sole cause.

Exam tips

• If a question only asks you to describe neural mechanisms, do not add evaluation. Examiners report that students waste time doing this.

• In a question on neural mechanisms, material about genes only earns credit if you link it to the brain. For example, explain how the MAOA gene affects serotonin.

• Strong evaluation of biological explanations includes animal evidence and extrapolation, cause and effect, determinism, reductionism and gene–environment interaction.

Possible Exam Questions 

1. Outline and evaluate the role of neural and/or hormonal mechanisms in aggression. (16 marks) 

2. Micah was experiencing outbursts of temper at home and school. His friends and family were concerned that he seemed to have no control over his temper. Micah has two older sisters, and they are both doing very well in school and are calm and easy going. Micah’s grandfather had spent some time in prison for arson and Micah’s father had died when Micah was younger in a fight. Using your knowledge of the genetic, neural and hormonal mechanisms in aggression explain why Micah may be having trouble controlling his temper. (12 marks)