Why changing your mind costs so much

Leon Festinger described, under the name of cognitive dissonance, the unpleasant state that appears when two elements of our thinking, or our thinking and our conduct, do not agree. This state pushes us to restore coherence, and the shortest route is rarely the one that consists in revising the most visible position.

One can reduce the tension otherwise: by minimising the awkward information, contesting its source, adding a reason that reconciles everything, or reformulating what one thought one had said. These exits are less costly than an admission, and that is precisely why they are the most frequent.

One detail of this work deserves to be known: the more costly the initial commitment was, the stronger the resistance to change. It is not attachment to the idea that makes revision difficult, it is what one has already invested in it.

A position expressed in front of others becomes a commitment. Revising it then requires not only giving up a conclusion, but doing so in front of witnesses, with the risk of being judged inconstant. The more forcefully and publicly the position was asserted, the higher that cost.

This mechanism is in no way irrational. Constancy is a real quality in social life, and a person who changes their mind at every argument heard becomes unpredictable, and therefore unreliable. The problem is not constancy, it is that it ends up being defended for its own sake, independently of what it protects.

In environments where positions are public and archived, this cost rises further: the record of what one has maintained remains consultable, and revision becomes an admission that can be used against one. That is a little-discussed effect of online discussion spaces.

Real changes of mind rarely take the form of a sudden reversal. They most often pass through a gradual reinterpretation: one discovers a distinction that makes it possible to keep one's principle while changing conclusion, or one shifts the subject of the disagreement.

This path is often invisible from outside, and it explains why people change more than they admit. Asking someone whether they have changed their mind produces a less reliable answer than comparing their positions at two different moments.

A few procedures reduce the bill. Explicitly distinguishing the objective from the means makes it possible to change one's mind about the means without giving up the objective, which is not a retreat. Announcing a position as provisional, backed by a precise piece of information, allows its revision when that information changes.

In a discussion, leaving the other an honourable way out is more effective than being right. A person backed into a corner protects their position even when they know it is fragile, because they are no longer defending an idea but their own seriousness.

The same gesture holds for oneself. Formulating in advance what would make you change your mind, and writing it down, turns future revision into the application of a rule you laid down, rather than a disavowal of yourself.

A profile is a photograph, not a verdict. Finding a different result years later is not a sign that one of the two measurements was wrong, but often that a revision took place without being noticed.

It is also why the positions most interesting to examine are not the firmest, but those where one hesitates: they are the only ones where new information still has a chance of weighing.