We would like to dispute in the strongest possible terms the theoretical underpinnings and proposed implications from Derren Brown’s ‘crowd experiment’ – The Gameshow – aired on Channel 4 on 28/10/11. Brown’s second instalment in his series of ‘experiments’ was designed to show us how being anonymous in a crowd can, in his words, “turn perfectly nice people into internet bullies, or rioters, or hooligans”. To demonstrate, audience members were led to believe they were participating in a new interactive game show in which the fate of an unwitting member of the public was placed in their hands. The ‘target’ was a young man who was out for a drink with some friends. Along with various actors, the man’s friends were in on the plan and were in contact with the studio via hidden earpieces.
Throughout the show, the audience were presented with a choice between two scenarios (one positive and one negative) for the man. The severity of the negative outcomes increased throughout the episode, and ranged from being mistakenly charged for an extra round of drinks, to being kidnapped by a ‘gang of thugs’. The audience chose the scenario with a negative outcome each time, and for Brown, this was evidence of the moral depravity that inevitably follows anonymity in crowds.
Whilst we welcome Brown’s efforts to popularise social psychology in innovative and engaging ways, this particular episode was premised upon outdated theory that led to misleading and dangerous conclusions. Before exploring these topics, it is worth briefly noting several methodological problems with the study. These include the fact that it was not actually an experiment (as claimed by the title) since no independent variable was manipulated (there was not a sample making equivalent decisions alone or without masks), the ‘bad’ choice was always presented to the audience second, the audience understood that the consequences of their actions weren’t ‘real’ (akin to an interactive episode of Beadle’s About!), and Brown – who offered the audience the choices – is renowned for his skill in influencing people’s decision-making processes. Whilst we take issue with these methodological confounds, the purpose of this piece is to question the psychological theory upon which the episode was based.
Brown stated during the episode and in a subsequent interview on his website that ‘deindividuation’ within crowds causes people to lose their identities and consequently behave in inevitably anti-social ways. Over thirty years of empirical work from the social identity tradition (for a review see Reicher, Spears, & Haslam, 2010) has discredited these claims. This research has shown that rather than a loss of identity within crowds, there is ashift from personal to social levels of identification. Instead of acting in terms of the norms and behavioural limits of one’s personal identity, within a psychological crowd one therefore acts in coherence with the norms of one’s salient collective identity. These norms will differ depending upon which social identity is salient at any given time, e.g. as a resident of a local community, supporter of a sports team, or as a member of an audience at a television recording. Crowd behaviour is therefore rooted in social context, such that individuals may even act more pro-socially in a crowd than they would do alone (see e.g. the non-violent resistance of Indian crowds in the face of colonial British rule, or within-crowd helping during emergencies [see Brown’s own blog on this topic]).
In the case of Friday’s ‘experiment’, the audience acted in terms of their collective identity as audience members in at least two ways. First, the very object of being in a game show audience is by definition to be entertained. Each time the audience were faced with a choice, they picked what was clearly the most entertaining option, and the selection that would prolong their involvement in the event. Second, the menacing masks that audience members wore were hardly neutral cues; in fact the very same masks were later worn by the ‘group of thugs’ who attempted the kidnap in the final scene. This is reminiscent of a famous study by Johnson and Downing (1979), who noted that when people were given robes resembling those of the Ku Klux Klan they displayed more anti-social behaviour than control participants. However, when participants were given nurses’ uniforms they displayed significantly less anti-social behaviour than controls. The fact that audience members wearing the ‘thug’ masks chose anti-social options is consistent with the argument that crowd behaviour is rooted in contextual cues, and not inherently anti-social.
It is important to emphasise that we are not arguing that crowds are immune from anti-social behaviour; some of the very worst atrocities in history have been committed by crowds (e.g. religious pogroms, lynchings etc.). The point is that crowd behaviour is neither intrinsically good nor bad, but is dependent upon the norms of the shared social identity of its members.
Why is all this important? Does it really matter to anyone other than social psychologists that outdated theory is portrayed as factual on prime-time television? The point is that an understanding of crowd psychology has important consequences for society. Regarding crowds as anti-social entities acting without identity or reason can legitimate their violent repression by security forces, prevent intragroup helping in emergencies, and facilitate the dismissal of popular protest as irrational by those in positions of power. Poor theory can therefore ultimately lead to both public disorder, and an attack upon our democratic rights as individuals to express our views collectively. It is therefore in all of our interests to gain an accurate understanding of crowd behaviour, rather than rely upon outdated theory that is not only wrong, but dangerous.
1. University of St Andrews
2. University of Sussex