When you do not know what you are feeling

Some people receive their states in blocks. There is something, it is unpleasant, it has been there since the start of the afternoon, and its exact nature does not appear. Others identify on the spot that it is irritation, more precisely irritation mixed with tiredness, and that the remark made at ten o'clock has much to do with it.

These two ways of functioning do not differ in the intensity of what is experienced, but in the resolution with which it is perceived. The first is not a degraded version of the second: it captures less detail and lets itself be less encumbered by analysis.

The term alexithymia was coined in 1973 by the psychiatrist Peter Sifneos, from the Greek: literally, the absence of words for emotions. He was describing patients who spoke abundantly about facts and symptoms and found themselves at a loss as soon as they were asked what they felt.

The notion was then applied to the general population, where it describes a characteristic spread along a continuum rather than a state apart. The available estimates put at around ten per cent the proportion of people scoring high on the scale published in 1994 to measure it, with important variations depending on the thresholds used.

Part of what is called not knowing what you feel is in reality a timing lag. The state is identified, but later: while recounting the day to someone, while writing, while thinking back on the scene two days afterwards, once the urgency has subsided.

This lag has a cost, that of reacting without knowing why, and a rarely named advantage: late identification is done cold, and therefore with less risk of mistaking the cause. Those who know straight away are also mistaken, and with more confidence.

It is often assumed that learning words would be enough. Research on what Lisa Feldman Barrett calls emotional granularity suggests otherwise: what distinguishes people is not the number of terms they know, but the use they make of them to distinguish close states in their daily experience.

One can have thirty words and use them interchangeably; one can use five in a perfectly discriminating way. Vocabulary lists posted in corporate corridors are therefore not enough, which does not make them useless objects: above all they make conversation possible.

A study published in 2007 by Matthew Lieberman and colleagues showed that naming the state visible on a face was accompanied by a drop in amygdala activity. The effect is small, obtained in a laboratory, over durations of a few seconds, and the participants questioned do not believe that naming helps them.

That last observation is the most interesting: the effect does not depend on the belief one puts into it. It does not make putting things into words a method, only an operation of which it can be said that it slightly modifies the course of what it describes.