The rule underneath the feeling
A child starts crying and an adult says, almost kindly, “Don't make it bigger than it is.”
Another child is angry and hears, “Anger helps you know when something matters.”
Someone grows up in a family where sadness is private, excitement is embarrassing and calmness is praised as evidence of being reasonable. Someone else learns that feelings are weather: real, temporary, informative, not sovereign.
Long before people study emotion regulation, they acquire theories about emotion.
Is emotion dangerous?
Useful?
Controllable?
Overwhelming?
A sign of truth?
A weakness?
Something that should be expressed, corrected, hidden, trusted, analyzed or endured?
These assumptions are emotion beliefs.
They matter because regulation does not begin from neutral ground. People regulate what they think emotion is.
Definition
Emotion beliefs are explicit or implicit assumptions about the nature, value, controllability, usefulness and consequences of emotions. Research has focused especially on whether emotions are seen as malleable or fixed, controllable or uncontrollable, and helpful or harmful.
Emotion beliefs are not the same as emotions themselves.
They are also not identical to emotion knowledge, emotional awareness, emotion regulation skill or cultural display rules, although all can interact.
A person may accurately identify fear while believing fear is unacceptable. Another may believe emotions are controllable but lack an effective strategy. A third may endorse emotions as useful while still avoiding their own.
Beliefs shape orientation toward emotional experience, but they do not fully determine what a person can do.
Fixed and malleable theories of emotion
Maya Tamir, Oliver John, Sanjay Srivastava and James Gross studied what they called implicit theories of emotion during the transition to college.
Some participants tended to see emotions as relatively fixed and difficult to change, an entity theory. Others viewed emotions as more malleable, an incremental theory.
Using a longitudinal and multimethod design, the researchers found that these theories were associated with emotional and social adjustment during a major life transition.
The work was influential because it framed beliefs as part of regulation rather than mere commentary on it.
If someone believes emotions cannot be changed, they may be less likely to invest effort in regulation. If they believe regulation is possible, they may be more willing to try.
But that logic has limits.
Believing an emotion is controllable does not guarantee that it is controllable in the desired way, at the desired speed, in the present context.
Optimism can support action. It can also create self-blame when emotion does not obey.
Ford and Gross: two belief dimensions
Brett Ford and James Gross's 2019 review organized much of the field around two broad belief questions.
First: are emotions good or bad, useful or harmful?
Second: are emotions controllable or uncontrollable?
These dimensions can shape how people approach regulation.
If anger is viewed as inherently bad, a person may regulate it quickly even when it contains information about a boundary violation. If sadness is viewed as useless, grief may become a problem to solve rather than an experience to integrate. If anxiety is viewed as uncontrollable, a person may not attempt regulation. If it is viewed as completely controllable, they may interpret persistent anxiety as personal failure.
The scientifically useful point is not that one belief is always correct.
It is that beliefs alter the regulatory landscape.
Beliefs can affect performance, not only preference
Bigman, Mauss, Gross and Tamir experimentally manipulated participants' expectations about emotion-regulation success.
Participants led to expect greater success subsequently showed greater regulation success during a negative emotion induction than control participants.
This is important because it moves beyond correlation.
Expectancy can influence what happens in a regulation attempt.
Yet the experiment does not prove that believing “I can control my emotions” is a universal intervention.
The task involved a specific induction, a particular regulatory context and short-term outcomes. Expectations may alter effort, attention, persistence, strategy use or interpretation of success. The mechanism is not reducible to positive thinking.
A bounded conclusion is stronger than a slogan: expected success can contribute causally to regulation performance under some experimental conditions.
Emotion beliefs are not always conscious sentences
People may not walk around thinking, “I endorse an incremental theory of affect.”
Beliefs can be embedded in habits, language and social response.
A person may say that emotions are healthy while apologizing every time they cry. A workplace may advertise psychological safety while rewarding emotional flatness in promotion decisions. A family may claim that anger is acceptable but become cold when someone expresses it.
This creates a measurement challenge.
Questionnaires capture endorsed beliefs.
Behavior may reveal different contingencies.
Social norms can influence what people are willing to endorse.
Emotion beliefs therefore sit across explicit cognition, learned expectation and cultural practice.
Beliefs about usefulness require emotional specificity
“Are emotions useful?” is too broad to answer well.
Fear can protect and distort. Anger can mobilize moral action and fuel aggression. Shame can signal social evaluation and become globally self-condemning. Pride can support persistence and slide into dominance. Sadness can orient people toward loss while narrowing motivation.
Even within one emotion category, usefulness depends on whether the emotion fits current reality, how intense it is, what action it motivates and how long it persists.
A belief such as “anger is bad” erases that functional diversity.
So does “all emotions are messages.”
Emotions carry information, but information is not infallibility.
An emotional state can reflect present conditions, prior learning, bodily state, expectations, social cues and interpretation.
Mature emotion beliefs have room for both respect and revision.
Controllability is not one thing
People can influence emotion in several ways.
They can change situations, attention, interpretation, bodily state, expression or behavior. They can also change the goal from reducing an emotion to carrying it differently.
A belief that “emotion is controllable” may therefore refer to different control targets.
Can I stop feeling this now?
Can I change its intensity?
Can I choose whether to act on it?
Can I recover more quickly?
Can I change the situation producing it?
Can I make room for it without escalation?
These questions should not be collapsed.
A person may have little immediate control over the first surge of grief and substantial control over whether they send a hostile message during it.
Belief measurement becomes more informative when the object of control is specified.
Beliefs interact with strategy choice
If someone sees emotions as dangerous and uncontrollable, avoidance may become especially attractive.
If they see emotions as malleable, they may invest more in reappraisal or problem solving.
If they see emotional expression as socially costly, suppression may become strategically rational even if it carries other costs.
This is why emotion beliefs belong beside emotion regulation flexibility.
Beliefs help determine which strategies appear available, legitimate and worth the effort.
But the arrow can run in the other direction too.
Repeated successful regulation may increase beliefs in controllability. Repeated failed attempts may reduce them. A coercive environment may teach that expression is dangerous. A supportive relationship may make previously intolerable affect more manageable.
Beliefs are shaped by experience as well as shaping it.
Culture is not a nuisance variable
Emotion beliefs are deeply cultural.
Communities differ in ideals about emotional control, expression, social harmony, authenticity, arousal and relational responsibility.
The 2025 cross-cultural meta-analysis of emotion regulation and mental health found that culture moderated associations between commonly studied regulatory strategies and mental-health outcomes.
That study is not a direct meta-analysis of emotion beliefs.
Its relevance is contextual: beliefs about which emotions should be expressed or controlled are embedded in the same cultural systems that shape strategy consequences.
A belief measure developed in North American English may not carry the same meaning in Swedish, Japanese, Arabic or multilingual contexts.
Translation requires conceptual validation, not only linguistic equivalence.
Development begins before theory can be articulated
Children learn emotion beliefs through response.
What happens when they cry?
What happens when they become angry at an adult?
Which feelings receive curiosity? Which trigger punishment? Which are laughed at? Which are rewarded because they make the child easy to manage?
Later, explicit beliefs may be layered over those contingencies.
A teenager may learn sophisticated language about emotional validation while still expecting rejection when they show need.
This gap between endorsed belief and learned expectation is an important frontier.
The current literature does not give us one developmental pathway from caregiver behavior to adult emotion beliefs. Family, peers, school, culture, temperament and life events interact.
Developmental claims should remain proportionate to evidence.
Organizations teach emotion beliefs every day
Institutions have unofficial emotion theories.
A hospital can treat patient fear as an inconvenience to workflow or as information requiring explanation. A newsroom can reward urgency until calm feels like laziness. A school can interpret anger as misconduct without asking what boundary, humiliation or injustice preceded it. A technology company can frame every negative user state as friction to be eliminated.
The Institute uses institutional emotion beliefs as a synthesis frame for these implicit rules.
It is not a validated organizational scale.
The frame asks which beliefs about emotion are embedded in policy, interface, incentive and professional culture.
Do systems treat emotion as noise?
As risk?
As data?
As something to monetize?
As a human response that sometimes deserves space rather than optimization?
These are design questions with ethical consequences.
AI can become an emotion-belief tutor without intending to
Conversational systems answer emotional questions constantly.
Every response carries assumptions.
“This feeling is understandable” teaches one thing.
“You should calm down before deciding” teaches another.
“This is your nervous system protecting you” can validate experience and also overgeneralize a mechanistic story.
“You can reframe this” may expand options and may imply that the problem is inside the interpretation when the environment is genuinely harmful.
AI systems therefore participate in the social transmission of emotion beliefs.
Current research has not established the long-term developmental effect of such interactions.
The Institute treats this as a priority research question: what emotion theories are embedded in systems that millions of people increasingly consult during emotionally salient moments?
Beliefs can conflict inside the same person
People do not necessarily hold one coherent theory of emotion.
Someone may believe sadness is useful in principle and unacceptable in themselves. They may encourage a friend to express anger while experiencing their own anger as dangerous. They may think anxiety is controllable at work and uncontrollable in intimacy.
This matters for measurement.
A global questionnaire asks for an average belief, but emotionally salient contexts can activate different learned rules. Explicit endorsement can also conflict with procedural expectation.
The person may sincerely say, “Feelings are welcome here,” while their body has learned that showing need leads to withdrawal.
Research on emotion beliefs therefore sits close to metacognition: not only what do I believe about emotion, but how accurately do I know which belief is operating now?
That second-order question remains less developed than broad mindset measures.
Beliefs are partly social forecasts
Many emotion beliefs are predictions about other people.
If I cry, they will think I am weak.
If I become angry, I will be abandoned.
If I show excitement, I will be embarrassed.
If I admit uncertainty, I will lose authority.
These are not beliefs about emotion in the abstract. They are beliefs about what emotion does inside a social field.
Sometimes the forecast is distorted by old learning. Sometimes it is accurate.
A workplace really may punish visible uncertainty. A family really may withdraw from anger. A marginalized person may face different social costs for the same expression than a high-status colleague.
This is why changing beliefs cannot be treated as a purely internal cognitive task.
Some beliefs need updating.
Some environments need changing.
Emotion beliefs can create self-fulfilling loops
If someone expects emotion regulation to fail, they may invest less effort, abandon strategies earlier or interpret normal fluctuation as evidence of helplessness.
If they expect success, they may persist longer or deploy a strategy with greater engagement.
Bigman and colleagues' experiment shows that expected success can influence performance under controlled conditions.
Across everyday life, such loops could accumulate, but that broader developmental claim needs direct longitudinal evidence.
The useful hypothesis is not “believe you can and you will.”
It is more precise: expectations can alter behavior inside regulation attempts, and repeated outcomes can then reinforce or revise expectations.
Belief and capacity may co-develop.
That is a stronger developmental model than either optimism or fatalism.
Beliefs can be accurate and still costly
Psychology often treats maladaptive beliefs as distorted beliefs.
Emotion beliefs complicate that assumption.
A person may accurately believe that visible anger will be punished in their workplace. A clinician in an emergency department may accurately believe that some emotional expression must be postponed until the immediate task is complete. A teenager may correctly predict that peers will mock vulnerability.
The cost does not make the belief false.
This is important because interventions aimed at “changing beliefs” can become invalidating if they ignore the social environment that made the belief sensible.
Sometimes the developmental task is revising an outdated forecast.
Sometimes it is finding a safer context.
Sometimes it is changing the institution so the belief no longer remains accurate.
Beliefs about emotion can conflict with beliefs about identity
Emotion beliefs also interact with identity rules.
“I should be able to control this” may be less about emotion than about being a competent leader. “I do not get angry” may protect an identity as kind. “I am an emotional person” may become a stable self-description that organizes what gets noticed and remembered.
This overlap creates another measurement boundary.
Researchers need to distinguish beliefs about how emotion works from beliefs about what kind of person one is allowed to be.
The constructs can reinforce one another without being identical.
What the evidence does not show
Beliefs can matter without functioning like commands. The evidence does not establish that:
- believing emotions are controllable makes every emotion controllable;
- fixed emotion beliefs are always harmful;
- incremental beliefs guarantee psychological health;
- viewing emotions as useful means acting on them is always wise;
- one questionnaire captures all emotion beliefs;
- expected success effects are equivalent to a general positive-thinking intervention;
- emotion beliefs are purely individual rather than socially and culturally learned;
- cultural differences can be reduced to simple national stereotypes;
- an organization's emotion beliefs can be inferred from its wellbeing statements alone;
- AI responses have already been shown to reshape users' emotion beliefs in a predictable long-term direction.
The evidence supports emotion beliefs as consequential. It does not support replacing emotional complexity with a new ideology about the “right” way to think about feelings.
Institute synthesis: every regulation strategy contains a theory of emotion
Before someone tries to change a feeling, they have already decided something about it.
Perhaps it is dangerous.
Perhaps it is temporary.
Perhaps it is evidence.
Perhaps it is noise.
Perhaps it is unacceptable in public.
Perhaps it is the truest thing in the room.
The Institute's synthesis is that regulation is never belief-free.
That is not a claim that every regulation act begins with a conscious proposition. It is a way of noticing the assumptions that make some actions feel obvious.
The next time emotion arrives, the most revealing question may not be “How do I regulate this?”
It may be the quieter one underneath:
What have I already decided this feeling means?
Research references
- Tamir M, John OP, Srivastava S, Gross JJ. Implicit theories of emotion: affective and social outcomes across a major life transition. Journal of Personality and Social Psychology. 2007;92(4):731-744. https://doi.org/10.1037/0022-3514.92.4.731
- Ford BQ, Gross JJ. Why Beliefs About Emotion Matter: An Emotion-Regulation Perspective. Current Directions in Psychological Science. 2019;28(1):74-81. https://doi.org/10.1177/0963721418806697
- Bigman YE, Mauss IB, Gross JJ, Tamir M. Yes I can: Expected success promotes actual success in emotion regulation. Cognition and Emotion. 2016;30(7):1380-1387. https://doi.org/10.1080/02699931.2015.1067188
- Chen MS, Cai Q, Omari D, et al. Emotion regulation and mental health across cultures: a systematic review and meta-analysis. Nature Human Behaviour. 2025;9:1176-1200. https://doi.org/10.1038/s41562-025-02168-8