Bad behaviour can spread quickly. Riots and uprisings – and violence and aggression more generally – can grow and proliferate in such a way that they’re often described as being “contagious”, like a disease. In fact, contagion is one the most persuasive metaphors for explaining collective behaviour.
The shared antagonism towards the police meant that these rivalries became superseded by a stronger group identity. The feeling of power this created among participants is what helped the riots grow and spread.
It seems to capture something so fundamental in human behaviour that it is invoked to explain both the spread of simple behaviours like smiling and yawning, and more complex interactions such as the spread of ideas in society or rapid changes in financial markets.
But despite the apparent success of this metaphor, research evidence suggests that such language actually conceals more than it reveals. As such, there are better ways to think about and explain the process of influence among large groups of people.
Specifically, new research my colleagues and I have carried out on the August 2011 riots in England suggests the violence did not spread mindlessly. Instead, we found that people were influenced by a sense of shared identity against the authorities, that transcended even the “postcode rivalries” of local gangs.
Our preliminary work on the riots that spread across English cities suggests social identity – rather than simple spontaneous contagion – shaped much of the behaviour that took place. First of all, not everyone in the affected cities joined in. Those that did shared an anti-police identity.
Second, many of those young people who did join in would more typically have seen each other as rivals, based on long-standing conflicts between different districts of the city.
We found that the shared antagonism towards the police meant that these rivalries became superseded by a stronger group identity. The feeling of power this created among participants is what helped the riots grow and spread.
This idea is in stark contrast to the typical description of crowd behaviour as contagious, which has long been used to undermine and attack the actions and motivation of large groups of people.
The police’s weak response encouraged even more action, helping the violence to spread to other targets. This was the case both for the original rioters and for others who heard about the events and then joined in.
The first use of the term “contagion” in a psychological sense was by the French historian Hippolyte Taine in his 1876 study of the French Republic. Taine borrowed from the terminology of medicine (not only “contagion”, but also “feverishness” and “delirium”) to capture what he saw as the barbaric mentality of the “mob”.
Plagiarising Taine’s ideas, the crowd theorist Gustave Le Bon (1895) defined contagion as uncritical passive social influence, arguing that any sentiment or behaviour could easily sweep through a crowd. He explicitly tried to use psychology as a weapon in a war against the working class “masses”, suggesting that if people lost their reason when they gathered in a crowd, there was no point reasoning with them and force was justified.
In this early usage, we can see the defining features of the contagion concept and some of its problems. First, comparing social influence to a disease implies that influence in crowds is something bad. Second, it suggests the spread is mindless, or even irrational, because it does not involve cognition.
Modern versions of the contagion concept have inherited these assumptions to varying degrees, meaning they can’t explain some important features of influence, both in crowds and between individuals.
In their 1969 critique of Le Bon’s account, the social psychologists Stanley Milgrim and Hans Toch pointed out that contagion couldn’t explain why riot police are impervious to the rousing effects of a demagogue, unlike others present in the same crowd. Similarly, Steve Reicher’s well-known study of the 1980 St Paul’s riot noted that while some behaviours did spread in the crowd (such as throwing stones at police) others did not (such as throwing stones at a bus).
A new model
Instead of a model of indiscriminate contagion to explain the way influence within and between crowds work, we need to use the notion of social identity, which means our definition of ourselves based on our membership of groups.
Our group membership might change from context to context. For example, we might define ourselves as an Arsenal supporter in a football context and as as a Christian in a religious context. This means different sources of influence can also vary according to the context.
Psychologists have shown that this social identity principle helps explain the spread of relatively simple behaviour, such as emotional responses. For example, one study involved showing a group of psychology students images of people displaying anger and fear. They were more likely to mimic the emotions if the people were described as other psychology students than if they were labelled as economics students.
The research my colleagues and I are conducting applies these ideas about social identity to a range of behaviours. Some types of simple behaviour often described contagious, such as yawning, really do seem automatic. Others (such as scratching) may depend more on the identity of the person the behaviour originates from. Our research will help to determine which behaviours are which.
Critics could point to the hundreds of experiments that provide evidence that some behaviour really is contagious, including relatively complex actions such as aggression. But the social identity research suggests that this may be because participants in the experiments share an implicit social identity with the target (such as “students”) and it is this that explains how they are influenced.
One of the features that distinguishes crowds from individuals is power. Our new theory is that the 2011 riots spread from location to location in part because participation gave individuals a greater feeling of empowerment. Shared identity led to greater expectations of support and these expectations of support led to action against the police.
The police’s weak response encouraged even more action, helping the violence to spread to other targets. This was the case both for the original rioters and for others who were not initially present but also identified against the police, who heard about the events and then joined in.
This kind of evidence suggests that we should abandon the idea of “contagion” and instead find different ways to talk and think about how people are influenced, in crowds and more generally.
There are other terms such as “spread” or “transmission” that more accurately describe the process. The advantage of using these more neutral terms is that we avoid misunderstanding the behaviour of people in crowds and treating it like a disease.