What your sense of belonging does to your opinion

Henri Tajfel's experiments on minimal groups showed that it is enough to divide people according to an insignificant criterion, sometimes drawn by lot, for them to start favouring members of their own group in the allocation of rewards. No shared history, no shared interest, no prior animosity is necessary.

This result sheds light on part of political functioning. Belonging does not need to be deep in order to steer judgements: the mere fact of perceiving oneself as being in a camp is enough to colour the evaluation of arguments, of people and sometimes of facts themselves.

It should be noted that the effect observed bears more on favour granted to one's own than on hostility towards others. That is an important qualification: the basic mechanism is favouritism, hostility coming afterwards, when real competition or a discourse that designates it is added.

A position may turn into a sign of belonging. When that happens, revising it no longer costs merely an admission of error, it costs a place in a group. One then better understands resistance to arguments that are otherwise solid: it is not the argument that is refused, it is the social price of changing.

The reverse also exists, less commented upon. A person may adopt a position because it is that of those around them, without having examined it, and discover on questioning themselves that they do not really hold it. Borrowed conviction is at least as frequent as defended conviction.

This shift from opinion to identity explains a troubling phenomenon: on certain subjects, the better informed a person is, the better they can defend their camp's position, without the probability of their changing it increasing. Competence then serves loyalty more than examination.

An opinion expressed in front of witnesses is no longer quite the same as an opinion held in private. Work on conformity shows that a share of people adjust what they say to what they believe to be the dominant view, without necessarily changing what they think.

This gap between the said and the thought has concrete consequences. It makes climates of opinion unstable: a position may seem to be in the majority in conversation and in the minority at the ballot box, or the reverse, simply because the cost of expressing it is not the same everywhere.

One useful signal is asymmetry of treatment: do you grant the benefit of the doubt to an argument depending on who makes it? Test it by imagining the same sentence spoken by someone from the other camp, and observe whether your demandingness changes.

Another signal is discomfort at voicing a reservation about a position of your own camp, when the same reservation would seem reasonable to you coming from elsewhere. That discomfort is a good indicator of the degree to which an opinion has turned into a belonging.

Answering a questionnaire without knowing what others answer neutralises part of these effects: there is neither a visible camp to join nor a gaze to satisfy. That is one of the points of an individual test, and the reason why it is better to take it before discussing it.

Order matters. A test that asks your position first and then tells you whom it resembles measures something different from a test that announces the camps in play from the outset. The first measures an opinion, the second also measures a wish to belong.