The Capgras Delusion and the Face That Looks Right but Feels Wrong

In 1923, the French psychiatrist Joseph Capgras and his junior colleague Jean Reboul-Lachaux published a paper on a 53-year-old woman they called Madame M. She insisted her husband had been swapped for a stranger who looked identical. So had her children. So had the neighbours. They named the syndrome l'illusion des sosies — the illusion of doubles.

For most of the twentieth century, Capgras delusion was filed under psychoanalysis: a defence against forbidden feelings, an unconscious split. Then in 1990, Hadyn Ellis and Andrew Young at the University of Wales proposed something stranger and more mechanical. Prosopagnosic patients — people with damage to the fusiform face area — cannot consciously recognise faces, but their skin still sweats faintly when shown someone they love. Recognition without feeling. Ellis and Young guessed Capgras was the same dissociation, run backwards. The conscious recognition pathway works. The autonomic one doesn't.

In 1997 Ellis and his colleagues tested it. Five patients with Capgras delusion were shown photographs of familiar and unfamiliar faces while their skin conductance was measured. Healthy controls spike when they see a face they know. The Capgras patients did not. Their fingertips stayed dry for everyone, friend or stranger.

That is the part the older theories couldn't reach. The face arrives at the visual cortex correctly identified — that's my wife — but the limbic chord that should sound underneath it is silent. The brain has to make sense of the gap. The conclusion it lands on is not that the feeling is broken. It's that the person is fake.

The delusion is, in a sense, the right answer to the wrong question.