Why Putting a Feeling Into Words Changes It

Introduction

Most people have had the experience of describing something to someone and finding that the description itself altered it.

Nothing was solved. No advice was offered. The situation was exactly as it had been ten minutes earlier. And yet the thing had changed shape slightly: become more bounded, more like an object and less like an atmosphere.

This observation has been studied under the name affect labelling, and it has produced one of the more interesting small findings in emotion research.

It has also produced a slogan that has travelled considerably further than the evidence behind it, which makes it a good case for looking at what was actually shown.

The Study That Made the Idea Famous

In 2007, a research group led by Matthew Lieberman scanned participants while they viewed photographs of emotional facial expressions.

In one condition, participants selected a word describing the emotion in the image. In another, they performed a task that required attending to the image without naming what it conveyed.

When participants labelled the emotion, activity in the amygdala, a structure heavily involved in processing emotionally significant material, was lower than when they did not. Activity in a region of the prefrontal cortex was higher, and the two were inversely related.

The interpretation offered was that translating an emotional experience into language engages a system that dampens the response, without the person intending to regulate anything.

The finding was striking, memorable, and widely reported.

Why That Result Needs Careful Handling

It is also the part of this literature that requires the most caution, for three reasons.

The sample was small, as was normal for imaging studies of that period. Small samples produce unstable estimates, and a great deal of early neuroimaging has proved harder to replicate than its initial reception suggested.

The second problem is more fundamental. Inferring a mental state from activity in a brain region is not straightforward, because regions are not dedicated to single functions. The amygdala is involved in fear, but also in processing salience, novelty and reward. Observing reduced amygdala activity does not establish that someone felt less afraid; it establishes that a region participating in many processes was less active. Reasoning backwards from brain activity to experience is a known inferential trap, and this literature is frequently reported as though it were not.

The third is simply that a difference in brain activity is not the same as a difference that matters to a person's life. Something has to connect the scanner to the world.

None of this means the finding is wrong. It means it cannot carry the weight often placed on it, and that better evidence has to come from somewhere else.

A Stronger Test, Outside the Scanner

That evidence arrived in 2012, from a study using a behavioural design rather than an imaging one.

Participants who were frightened of spiders approached a live tarantula across several sessions. They were assigned to different conditions: some described their emotional reaction in words as they went, others reframed the situation in less threatening terms, others were distracted, and others simply did the task.

At follow-up a week later, the group who had put their feelings into words showed a greater reduction in physiological arousal and were able to approach more closely than the others.

This is a considerably more persuasive result than the imaging study. It has a behavioural outcome, a delay before measurement, and active comparison conditions rather than a passive control.

It also connects to something established elsewhere in this field, that what produces durable change when approaching a feared situation is not how calm the person manages to feel at the time.

The Odd Part, It Does Not Feel Like It Is Working

Here the research becomes genuinely counterintuitive.

Across studies, participants generally do not report that labelling makes them feel better in the moment. In some cases they report feeling somewhat worse. And when asked in advance which strategy they expect to help, people tend to rank affect labelling low, often below distraction, which performs less well.

This has led researchers to describe affect labelling as an implicit form of emotion regulation: something that regulates without being experienced as regulation.

The practical consequence is awkward. A strategy that provides no immediate sense of relief is one almost nobody will persist with voluntarily, because the feedback signal points the wrong way. It resembles the situation with approaching feared situations, where the felt experience during the attempt is a poor guide to whether anything useful occurred.

It does not feel like relief. That may be why so few people persist with it long enough to notice it worked.

It should also be said that findings on self-reported emotion have been less consistent than those on physiological and neural measures. This is a real gap in the literature rather than a detail, a strategy that reliably alters physiology while leaving subjective experience untouched is not obviously the same thing as one that helps.

Having More Words for It

A related line of research concerns not whether people label their emotions but how finely.

Some people, asked how they feel, distinguish between irritated, resentful, humiliated and let down. Others report the whole territory as bad. This capacity to make fine distinctions has been termed emotion differentiation, or granularity.

Studies have found that people higher in differentiation tend to draw on a wider range of regulation strategies, particularly when distress is high, and that lower differentiation is associated with poorer outcomes across several difficulties.

The proposed explanation has a certain logic. Bad does not indicate a course of action. Let down by someone I relied on does. A more precise description narrows the field of sensible responses, where an imprecise one leaves only the general instruction to feel better.

This literature is largely correlational, and the direction of the relationship is not settled, distress may reduce the capacity for fine distinctions rather than the reverse.

What the Evidence Actually Supports

The popular version of all this is the phrase name it to tame it, which is memorable and overstates the case in two directions at once.

It implies the effect is large, when it is modest. And it implies the effect is felt, when the research suggests it largely is not.

What the evidence supports is narrower and still worth having.

Describing an emotional experience in specific language appears to change how it is processed, in ways detectable physiologically and behaviourally, and with effects that show up later rather than immediately. Doing so precisely appears to be more useful than doing so vaguely. And the absence of any felt improvement at the time is not evidence that nothing happened, which is worth knowing, because that absence is exactly what would otherwise lead someone to abandon it.

None of which makes this a technique that resolves distress. It is a small effect from a research literature with real limitations, and it is not a substitute for support with anything serious.

But it does supply a reason for something people often do instinctively and then dismiss as pointless: saying the thing out loud, in words, to someone. The instinct appears to be better founded than the immediate feedback suggests.

If difficult emotions are a persistent feature of your life, that is worth raising with a GP or a qualified professional rather than addressing alone.