What Feelings Are For

The number moved with the person

In the 1990s, while I was studying exercise physiology, I tested athletes, and I kept running into something the textbooks had not prepared me for. The numbers moved with the person’s state. Same athlete, same training, different day, different result, and nothing in the training log explained it. I finished the degree in 1998, coached rowers, and then spent most of a career using exercise to help people manage health and disease. The puzzle came along with me. A person’s emotional interior mattered to how they recovered as much as the medicine and the exercises did, and I had no good way to say why.

This is the first of four pieces that try to say why, and it starts with the question underneath the others. What is a feeling for? Most advice treats emotion as weather around the work. The research I find most convincing says emotion is one of the ways the brain decides what the work is: what deserves attention and effort, and what happens next. That makes the goal something other than calm. It is to understand each feeling, use what it offers, and keep any single one from running the whole house.

I should say plainly what kind of claim that is. It is a claim about the system, not about every episode. A feeling can be the wrong size, at the wrong time, or about the wrong thing, and when anxiety or low mood is intense, persistent or disabling it is a matter for a clinician, not a signal to be mined. I am a reader of this literature, same as you, and I am still learning it. Where the evidence is thin, I will say so.

First, what we are talking about

Words in this field get used loosely, and I have caught myself doing it. These are the working definitions I will hold to across the series.

What is an emotion? A short tour

People have argued about this since the nineteenth century, and your answer changes what you think anyone should do about a feeling. If a feeling is just the body’s reaction, you manage the body. If it is a judgment, you examine the judgment. The timeline below is simplified: first the body, then evaluation, then a brain that builds the experience in the moment.

The most recent is Lisa Feldman Barrett’s theory of constructed emotion: the brain predicts what the body needs next, using signals from inside the body (interoception) and learned concepts, and an emotion is the result.¹ It is contested, and nothing here depends on it being the last word. One of its predictions has been tested hard, though, and it matters to anyone who performs, coaches or teaches.

The prediction is that a named emotion has no single bodily signature. A 2018 meta-analysis pooled 202 laboratory studies of autonomic responses (heart, breathing, sweat) during induced emotion and found large variation within each category and real overlap between them, not a distinct fingerprint for fear, anger or sadness.² A racing heart does not announce which feeling it belongs to. The name is a judgment laid over arousal, and judgments can change. So how a person performs depends on two things at once: how activated the body is, and what the mind makes of it.

Anxiety and excitement have nearly the same arousal. The difference is in the reading.On the same heartbeat

Two dimensions that carry most of the weight

A useful map places feelings on two axes, how pleasant and how activated. Anxiety and excitement sit in the same half of it because both are highly activated.

The map also helps with a creativity result. A meta-analysis of 102 effect sizes found that positive moods improved creativity over neutral ones, carried mainly by activating positive moods such as happiness; relaxed positive states did not do the same. Fear and anxiety were associated with lower creativity, particularly cognitive flexibility, while sadness showed no clear effect.³ The pooled positive-mood effect was small (r = .15), but calm was not the state that produced the best new ideas. I did not expect that, and it is one reason I distrust any plan that treats calm as the universal target.

Emotion as information

Imagine running a day without it. Every task would weigh the same, and nothing would pull you toward the report that is due or away from the ledge. A system with no way to say this matters, this is risky, this is new, hurry has to compute everything at full cost, and emotion is a cheap way of making those calls. Most researchers now see it that way: not a storm that passes through the mind, but one of the ways the mind decides what to spend itself on.

A working model helps me think about it. What a person produces under pressure depends on at least four things that influence one another: the capacity available (skill, knowledge, working memory), how the situation is appraised and felt, where attention is pointed, and how activated the body is. The task sets the demands and the stakes, and what comes out, along with what it cost, flows back into the loop. The diagram is conceptual. Nothing in it is a number, and the links are not multiplication or any formula. It shows that the parts shape each other, and that a shortfall in one can be felt in all the rest.

The model does not push emotion off to the side as a distraction. Appraisal is one of its four parts, and appraisal is how a situation gets its weight. An athlete who feels nothing before a final, a surgeon who feels nothing about the operation, a student who feels nothing about the exam: each is missing a source of the energy that makes effort worth spending. The trouble is feeling that is misdirected, or more than the task can use.

Positive feeling is not a free pass

It is tempting to answer the dangers of anxiety with the benefits of joy. The broaden-and-build theory proposes that positive emotions widen thinking, which builds resources, which feeds more positive emotion. A 2024 network analysis found that the widening step, the part that makes the theory distinctive, sat largely apart from the rest.⁵ Overselling positivity costs the same as overselling calm: it turns a finding into a mood you are supposed to produce on command.

What each feeling is for

Poets got there first. The thirteenth-century Persian poet Rumi is best known in English for a short poem, in the Coleman Barks rendering called “The Guest House,” that pictures a person as a house where joy, sadness, meanness and every other feeling arrive unannounced as visitors, all to be welcomed. That translation is under copyright, so I am paraphrasing, and I would send you to the poem itself. It holds the stance of this series: not a feeling to be ejected, but a visitor with a job. It is a poem, not a study, and the research below carries the weight.

The research tradition behind the stance is real, too. Functional accounts of emotion, set out in a 1999 paper by Dacher Keltner and James Gross, treat emotions as solutions to problems of physical and social survival: coordinated bundles of experience, body response and action that tend to help. They look for function at three levels. Inside a person, emotions coordinate the body, perception and thinking. Between people, embarrassment appeases and an infant’s distress draws care. Across a society, they help hold groups together.⁶ Two qualifications: this is a theory about the system, not every episode, and intense or persistent anxiety or low mood is a matter for care.

Fear and anger are not the same bin

The comparison breaks a habit of thought. A program of “reducing negative emotion” would put fear and anger in one pile, but they do not behave alike. In a set of experiments, fear raised perceived risk and pushed people toward the safer option, while anger lowered perceived risk and pushed them toward the riskier one. The authors traced the difference to appraisal: anger comes with a sense of certainty and control, fear with uncertainty and a lack of control.⁷

Sadness, and what a bad mood can do

A review of the benefits of negative mood reports that it can improve memory, reduce certain judgment errors and boost motivation in some settings.⁹ That does not make low mood a performance aid, but it does mean “bad feeling” holds more than one thing, and that wanting to be rid of all of it is a different goal than understanding it.

Regulating for use, not for comfort

If emotions have functions, what we are trying to do when we regulate them changes. Maya Tamir’s taxonomy separates hedonic motives, about how a feeling feels, from instrumental ones, about what else the feeling can give us: behavioral, epistemic, social and eudaimonic benefits.¹¹ People sometimes want to feel something unpleasant because it serves a goal, and sometimes they are right.⁸ That is a more grown-up account of regulation than a drive toward comfort, and I find it a relief.

The real danger is takeover

Under high, uncontrollable stress, catecholamines weaken prefrontal control and strengthen older, reflexive systems.¹² The trouble is not that the emotion exists but that it has become the only channel, and the plan, the cue, the next action and the other person get crowded out. The remedy is not silence. It is a system with more than one voice in it. Part 2 is about how that crowding-out happens.

When the feeling is telling the truth

The interference story, in which anxiety eats the mental resources a test needs, is intuitive and popular. It is not always right, and the best counterexample I know of changes what you do first.

Consider 309 German medical students preparing for their final state exam. Researchers measured trait test anxiety, tracked performance across a 72-day preparation phase, and compared it with mock-exam and final-exam results. Test anxiety did correlate with lower exam scores (about r = −.20 for the final). But it did not predict a drop from mock exam to final exam, and it no longer predicted the final score once the mock-exam score or preparation performance was controlled for. Anxious students had learned less during preparation. Daily anxiety did not predict fewer correct answers that day, while fewer correct answers did predict more anxiety the next day.¹³

That is strong evidence against interference in that sample, and the authors say so: the sample was selective and high-achieving, the students practiced with exam-style questions, and interference may still appear in novel test situations. The lesson is not that anxiety does not matter. It is that the arrow can run the other way. Anxiety may be a reasonable response to knowing you are less ready than you would like to be, and some of what looks like a feeling problem is a preparation problem the feeling is reporting accurately.

Anxiety also has a well-described way of hiding: people can compensate with extra effort until the effort runs out, so the anxious person hitting the numbers may be paying more than anyone sees.^4,14 Part 2 takes that up in full. The point for now is that the feeling can be accurate, or loud, or both.

I find the sorting question freeing, because it takes some of the shame out of a bad feeling. If the alarm is right, act on it. If the alarm is too loud for the danger, turn it down. In both cases the feeling is doing its job. The mistake is reaching for a tool before asking which case you are in.

Learning to listen

The first part of any practice is a way of listening, not a technique. Here are four moves I would start with. They are my own synthesis of the functional and appraisal research, a hypothesis and not a validated protocol, and I would not make them heavier than they are. Don’t over complicate, just start.

Keeping notes on the driver

If feelings are information, the next question is whether I can read them any better than I did last year. I think the answer starts with writing them down. Most of what we track is the engine: pace, load, hours, scores. Very few of us ever log the driver. A line or two after a hard session, an exam or a difficult meeting is enough: what I felt, what I thought it was about, what I did, and how it went. Then, a few weeks on, go back and check my guesses against what happened. Guess, then check. I have been wrong about my own state often enough that this is the only habit that has made me better at reading it.

There is a second thing I learned the slow way. Notes about your inner life are different from notes about your split times. The moment I suspect someone else may read them, I start writing for that reader, and the notes stop being true. So I think a person should decide who sees what, and in how much detail. That is not secrecy. It is the condition under which the log stays honest.

A log you edit for your audience is no longer a log of you.On private notes

Nothing here is a measured result. It is how I have come to think about it, and the evidence that keeping such a log improves performance or wellbeing is not something I can cite. It is a habit to test on yourself, and the test is simple: do your guesses get better?

What this does not show

Before the turn, some limits, in plain terms.

• This is a lens, not a diagnosis. The functional view of emotion is a theory about the system. Individual episodes can be mismatched, and I am not telling anyone that a bad feeling is secretly good.
• Persistent anxiety or low mood deserves a clinician. If it is intense, lasts, or gets in the way of daily life, that is a matter for care, not for an essay.
• The sorting question and the systems diagram are my own synthesis. They follow from the cited work, but they are hypotheses, not validated tools, and the figures illustrate them.
• The medical-student study is one sample. The authors caution that it was selective and high-achieving and that interference may appear elsewhere. I use it to show that the arrow can run backward, not to claim anxiety never interferes.
• The medical-student scene is constructed. It is labeled as such and does not report a real person.
• The note-keeping habit is untested. I describe what I do and why. I am not citing a trial that it works.