The post
A Korean-language Threads post, at about 40,600 views when we read it on November 16, 2026, says that under hundreds of feelings there are only four desires: the wish to be loved, to be safe, to be recognised and to be free. The second post of the thread pairs each wish with the feeling that signals it. Hurt at being let down (서운함) points to the wish to be loved. Anxiety (불안) points to the wish to be safe. A sense of being wronged (억울함) points to the wish to be recognised. Irritation, or feeling stifled (짜증·답답함), points to the wish to be free. The glosses are ours and approximate.
It then gives a method. Do not get rid of a feeling or avoid it. Ask what it really wants, because a feeling is a signal of what your body wants. A feeling you avoid or press down does not vanish. It stays in places around the body.
That is three claims: a count, a map from feelings to wishes, and a method with a warning. Each can be set against research.
How many desires are there?
Psychologists have listed basic needs for decades and have not settled on a number.
| List | What it names | Count |
|---|---|---|
| Ryan and Deci, 2000 | competence, autonomy, relatedness | 3 |
| Sheldon and colleagues, 2001 | ten tested; self-esteem, relatedness, autonomy and competence stood out | 4 of 10 |
| Tay and Diener, 2011 | basic needs, safety, social support and love, respect and pride, mastery, autonomy | 6 |
| Kenrick and colleagues, 2010 | physiological needs, self-protection, affiliation, status and esteem, mate acquisition, mate retention, parenting | 7 |
| Reiss, 2004 | basic desires that are largely unrelated to each other | 16 |
The post’s four fit inside these lists. Being loved sits with relatedness, affiliation, and social support and love. Being recognised sits with esteem, respect and status. Being free sits with autonomy. Being safe sits with safety and self-protection. What the post leaves out is competence, or mastery, which three of the first four lists name. The evolutionary list adds mating and parenting, and has no separate motive for freedom.
Safety is the contested one. In Sheldon’s study of what makes everyday events satisfying, security ranked eighth of ten needs. In the events that went badly, though, a lack of security stood out: people wished for the safety and predictability they usually take for granted. Tay and Diener found safety a relatively weak predictor of well-being across 123 countries, while Connolly and Lindh, working with five needs, found only autonomy and security still significant in their models. Safety seems to be noticed when it is missing, which suits a feeling like anxiety that rises on threat.
Culture reorders the list. In Sheldon’s South Korean sample of 200 students, relatedness topped the needs in satisfying events and self-esteem came second. In data from 66 countries and nearly 100,000 people, autonomy went with well-being across the world, and more strongly in wealthier and more individualistic countries.
“Hundreds” is a count of words. A study in which people reported what 2,185 short videos made them feel found 27 kinds of emotion, joined by gradients rather than sharp edges, and noted that people describe emotional experience with hundreds, if not thousands, of terms. Across 2,474 languages, which feelings share a single word varies, and the variation follows where language families live. What the language families share is how pleasant a feeling is and how much it activates the body. A short list is a summary someone chose, not a bottom layer found under the feelings.
Does a feeling point to a wish?
Some pairings have support. Hurt follows events that tell a person they are valued less, usually from someone they know well, in a study of 164 people. Mark Leary’s review gathers seven feelings that arise when a person’s value to others seems low or at risk, among them hurt, loneliness, shame and jealousy. That fits hurt with the wish to be loved. In other studies the same devaluation brought both hurt and anger, but hurt carried the goal of restoring the other person’s acceptance and anger the goal of changing their behaviour (Lemay and colleagues). In an online experiment with 739 people, failing an exam caused social pain much like being rejected, and the authors trace it to a blow to self-esteem, which bears on the wish to be recognised. In imagined scenarios, Smith and Lazarus found support for links between how a person sizes up a situation and feeling fear or anxiety, anger or guilt, and weaker support for sadness. Kenrick and colleagues list unfamiliar places, darkness and unfamiliar angry men as the triggers of self-protection, whose active states are anxiety and fear. Anxiety and the wish to be safe fit well.
Other pairings run one way. In two studies where messages urged students to floss or to drink less, a sense that the message threatened their freedom led to anger and negative thoughts, and the effect on attitude and intention ran through them. That fits freedom with irritation. It does not show the reverse, that irritation means a wish for freedom, and anger is common elsewhere. Across 37 countries and 2,921 students, anger events were the ones most often judged very unfair, followed by disgust, sadness, fear, guilt and shame, and judging an event unfair had strong effects on every negative feeling studied. A sense of being wronged is a family of feelings, and its wish could be fairness as easily as recognition.
So the relation is not a lookup table. One wish sits under several feelings, one feeling can serve several wishes, and one situation can bring several feelings at once. It is a loose map, strong in a few places.
Asking what a feeling wants also depends on who is asking. Norbert Schwarz’s account of feelings as information holds that what people conclude from a feeling depends on the question they bring, and that they stop relying on a feeling once they attribute it to something else. Nisbett and Wilson’s classic review of experiments showed that people explain their own responses with causal theories rather than by looking inward. An answer to “what do I really want?” can be a good lead. It is also a story the person holds about themselves.
The body part of the claim has more support. When 701 people coloured the places on a body outline where they felt different emotions, each emotion had its own map, and the maps were alike in West European and East Asian samples. A later study of 3,954 people from 101 countries found the same. People report feeling emotions in consistent places in the body. That does not show a feeling carries a message about what the body wants.
Ask, or press it down?
The method has two halves, and they fare differently.
Attending helps, in some forms. Putting a feeling into words reduced the amygdala’s response to emotional images in a brain-imaging study. When distressed, people who tell their feelings apart finely are less likely to turn to binge drinking, aggression or self-injury, in Kashdan, Barrett and McKnight’s review. Asking “why” from a distance, as an observer, let people think through an anger-provoking memory without feeling the anger again, in two experiments by Kross and colleagues. A recent meta-analysis of 13 trials found a therapy built on working with feelings, emotion-focused therapy, beat inactive control groups by a wide margin (g = 1.39) but not other psychotherapies (g = 0.28, range −0.19 to 0.76), and its authors call most of the trials poor in quality and note signs of publication bias. Most of this tests naming a feeling and asking why, not the post’s own question, what a feeling wants.
Venting and dwelling do not. In Bushman’s experiment, angered people who hit a punching bag while thinking about the person who had angered them felt angrier and were more aggressive than people who did nothing, and doing nothing beat venting. In a review of 114 studies, rumination had a large link with symptoms of anxiety, depression, eating and substance disorders, against medium to large for suppression and avoidance (Aldao and colleagues). Asking is not the same as going over it again.
Pressing a feeling down is several different things. In a meta-analysis of 306 comparisons, hiding the outward sign of a feeling changed measured emotional outcomes a little (d = 0.32), while trying to press down the experience itself or the thoughts did not (−0.04 and −0.12). Distraction worked (0.27) and concentrating on the feeling did not (−0.26). Habit matters too: in seven samples from Taiwan and the US, over 1,300 people in all, habitually hiding positive emotions predicted lower well-being, and habitually hiding negative ones did not (Yu and colleagues).
Does it stay in the body? Measured, that means several things. In a film study of 85 people, those told to hide their reaction to a disgusting film showed less on their faces, felt no less, and showed signs of more sympathetic nervous activity. Later work disputes the middle point: in two within-person experiments, suppression did lower reported negative emotion (Jones and colleagues). The arousal holds up better. A 2023 meta-analysis of 24 studies found manipulated suppression raised stress-related physiology a little (g = 0.20), though self-reported habitual suppression did not, overall. Hiding feelings left poorer memory for the event, and in conversations raised the blood pressure of the person on the other side. Over years, in 729 US adults followed for 12 years, those who scored higher on a suppression scale were more likely to have died, by a margin that only just reached significance (hazard ratio 1.35 for the 75th against the 25th percentile, 95% interval 1.00 to 1.82). The ratio was 1.70 for cancer deaths, which also only just reached significance, and 1.47 for cardiovascular deaths, which did not. A meta-analysis of 22 studies and 6,775 people found repressive coping linked to cancer and high blood pressure but not to cardiovascular disease in general, heart attack or asthma. These are costs, some of them bodily. None measures a feeling stored in a place.
The cost varies by culture. In one comparison of 71 European American and 100 Hong Kong Chinese students, holding back expression went with poorer psychological functioning only among the European Americans. In another, the social costs of suppression were smaller for people with more Asian values. In a 14-day diary study, people with a more interdependent view of themselves gained well-being and relationship quality from holding back negative feelings while making a sacrifice for a partner, and less interdependent people lost both (Le and Impett). Korea has its own named syndrome for anger held in, hwabyung. A proposal for a new anger disorder describes it as a chronic anger syndrome of Koreans, marked by pent-up anger and bodily symptoms, and a review of seven qualitative studies of patients calls it known to occur through the continued repression of anger. Those are patients’ accounts and a diagnostic proposal, not a test of cause.
What holds up
- “Four desires sit under hundreds of feelings.” Not as a count. Lists of needs run from 3 to 16. “Hundreds” counts words, and languages cut them differently. The four do overlap the lists.
- “Each feeling points to one wish.” Partly. Hurt and anxiety fit their wishes. Irritation runs one way, and being wronged spreads over many feelings.
- “Ask what the feeling wants. Do not press it down.” Mostly, with conditions. Naming a feeling and a distanced “why” help. Venting and dwelling do not. Hiding a feeling has costs that vary by culture and by what is hidden. The studies above do not test whether a feeling reveals a wish.
What survives is smaller than the post and still useful. Feelings follow how a person sizes up a situation against what matters to them, and the same few wants come up across the lists. Which wants, how many, and which feeling signals which differ by list, by language and by person.
What it means for a framework
A framework is a written account of how one person reasons. These studies are of people, not of frameworks, so what follows is what the evidence supports about one part of it, what a person is reaching for.
Reasoning bends toward what a person wants and fears. Kunda’s review found that people are more likely to arrive at the conclusions they want to arrive at, held back only by the justifications they can build. What a person wants shapes which conclusions they reach, so it belongs in a record of how they reason.
A short shared list, with weights that belong to the person. The lists agree more than they differ on love or belonging, safety, esteem and autonomy, and add competence. The post’s four plus competence would cover most of them. What differs between people is how strongly each pulls, what sets it off, and what they do when it is threatened. Kenrick and colleagues tabulate triggers and individual differences for each of their motives, and Reiss’s 16 desires are largely unrelated to each other, so a person looks more like a profile than a type. Culture moves the weights too, as in the Korean students’ ordering above.
Learn it from what the person chooses and how they react, not from what they say they want. A meta-analysis of 49 papers and 6,151 people found that implicit measures, which infer a motive without asking about it, correlated only 0.13 with questionnaires asking what people want. Nisbett and Wilson’s finding points the same way. The lab’s own tests show a version of it for taste: a selector that learns from a person’s own ratings of jokes did 2.9 points better than a shared one after 20 ratings and 6.4 points better after 80 (A selector that learns one person).
Keep it short and grounded in what the person did. Three 2026 preprints agree on a direction. In SCOPE, built from a two-hour protocol answered by 124 people, demographics explained about 1.5% of the variance in how similar people’s responses were, personas built from traits and identity narratives matched or beat the best full-information case on correlation with human answers, and model-written summaries, even long ones, never beat the grounded profiles. A second paper found that more prompt-driven reasoning worsened value polarisation, and that a value check trained on real human data improved fidelity. A third found that richer persona descriptions reduced diversity, and that simple age-and-gender personas beat richly specified customer profiles. A record of wants should be brief, tied to real responses, and not embroidered by a model.
For people known only from their record, mark what is inferred. A framework of someone who cannot be asked has their stated aims and what they did, not their own account of what they were after. Models are weak at the step from situation to feeling to wish: Yeo and Jaidka found they can reason from context to emotion and back “to a certain extent”, but are poor at associating situational outcomes and appraisals with specific emotions. A wish read off a public figure’s record is an inference, and belongs apart from what the person said.
The lab’s test of five layouts of one designer’s record, scored on his 56 later projects, shows where not to look. Every framework layout hit his real central move 12 to 21% of the time, and his name alone hit 26% (The shape of a framework). A record of what a person protects is a change of content, not of layout. In one line: record what the person’s reasoning protects, short, and learn it from what they did.
Sources
- Threads post by emotion_lab_bium, in Korean, and its second post in the same thread, read November 16, 2026, about 40,600 views.
- Cowen and Keltner, Self-report captures 27 distinct categories of emotion bridged by continuous gradients, Proceedings of the National Academy of Sciences 114(38), 2017.
- Jackson and colleagues, Emotion semantics show both cultural variation and universal structure, Science 366, 2019.
- Ryan and Deci, Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being, American Psychologist 55(1), 2000.
- Sheldon, Elliot, Kim and Kasser, What is satisfying about satisfying events? Testing 10 candidate psychological needs, Journal of Personality and Social Psychology 80(2), 2001.
- Tay and Diener, Needs and subjective well-being around the world, Journal of Personality and Social Psychology 101(2), 2011.
- Kenrick, Griskevicius, Neuberg and Schaller, Renovating the Pyramid of Needs, Perspectives on Psychological Science 5(3), 2010.
- Reiss, Multifaceted Nature of Intrinsic Motivation: The Theory of 16 Basic Desires, Review of General Psychology 8(3), 2004.
- Fors Connolly and Lindh, Socioeconomic Status, Need Fulfillment, and Subjective Well-Being, Journal of Happiness Studies 26(6), 2025.
- Martela, Joshanloo and Krys, Autonomy is Associated with Well-being Across the World, but more Strongly in Wealthy and Individualistic Countries, Social Indicators Research 181, 2025.
- Smith and Lazarus, Appraisal components, core relational themes, and the emotions, Cognition & Emotion 7(3-4), 1993.
- Leary, Springer, Negel, Ansell and Evans, The causes, phenomenology, and consequences of hurt feelings, Journal of Personality and Social Psychology 74(5), 1998.
- Leary, Emotional responses to interpersonal rejection, Dialogues in Clinical Neuroscience 17(4), 2015.
- Lemay, Overall and Clark, Experiences and interpersonal consequences of hurt feelings and anger, Journal of Personality and Social Psychology 103(6), 2012.
- Hudd and Moscovitch, Social pain and the role of imagined social consequences: Why personal adverse experiences elicit social pain, with or without explicit relational devaluation, Journal of Experimental Social Psychology 95, 2021.
- Dillard and Shen, On the Nature of Reactance and its Role in Persuasive Health Communication, Communication Monographs 72(2), 2005.
- Mikula, Scherer and Athenstaedt, The Role of Injustice in the Elicitation of Differential Emotional Reactions, Personality and Social Psychology Bulletin 24(7), 1998.
- Schwarz, Feelings-as-Information Theory, in Handbook of Theories of Social Psychology, volume 1, 2012.
- Nisbett and Wilson, Telling more than we can know: Verbal reports on mental processes, Psychological Review 84(3), 1977.
- Nummenmaa, Glerean, Hari and Hietanen, Bodily maps of emotions, Proceedings of the National Academy of Sciences 111(2), 2014.
- Volynets, Glerean, Hietanen, Hari and Nummenmaa, Bodily maps of emotions are culturally universal, Emotion 20(7), 2020.
- Lieberman and colleagues, Putting Feelings Into Words, Psychological Science 18(5), 2007.
- Kashdan, Barrett and McKnight, Unpacking Emotion Differentiation, Current Directions in Psychological Science 24(1), 2015.
- Kross, Ayduk and Mischel, When Asking “Why” Does Not Hurt: Distinguishing Rumination From Reflective Processing of Negative Emotions, Psychological Science 16(9), 2005.
- Whittington, Jafari and Borgogna, Emotion-focused therapy for distress-based psychopathology in adults: A systematic review and meta-analysis, Psychotherapy Research, 2026.
- Bushman, Does Venting Anger Feed or Extinguish the Flame? Catharsis, Rumination, Distraction, Anger, and Aggressive Responding, Personality and Social Psychology Bulletin 28(6), 2002.
- Aldao, Nolen-Hoeksema and Schweizer, Emotion-regulation strategies across psychopathology: A meta-analytic review, Clinical Psychology Review 30(2), 2010.
- Webb, Miles and Sheeran, Dealing with feeling: A meta-analysis of the effectiveness of strategies derived from the process model of emotion regulation, Psychological Bulletin 138(4), 2012.
- Yu, Haase and Chang, Habitual Expressive Suppression of Positive, but not Negative, Emotions Consistently Predicts Lower Well-being across Two Culturally Distinct Regions, Affective Science 4(4), 2023.
- Gross and Levenson, Emotional suppression: Physiology, self-report, and expressive behavior, Journal of Personality and Social Psychology 64(6), 1993.
- Jones, Isaacowitz and Ayduk, Conceal and Don’t Feel as Much? Experiential Effects of Expressive Suppression, Personality and Social Psychology Bulletin, online 2024.
- Tyra, Fergus and Ginty, Emotion suppression and acute physiological responses to stress in healthy populations: a quantitative review of experimental and correlational investigations, Health Psychology Review, online 2023.
- Richards and Gross, Emotion regulation and memory: The cognitive costs of keeping one’s cool, Journal of Personality and Social Psychology 79(3), 2000.
- Butler, Egloff, Wilhelm, Smith, Erickson and Gross, The social consequences of expressive suppression, Emotion 3(1), 2003.
- Chapman, Fiscella, Kawachi, Duberstein and Muennig, Emotion suppression and mortality risk over a 12-year follow-up, Journal of Psychosomatic Research 75(4), 2013.
- Mund and Mitte, The costs of repression: A meta-analysis on the relation between repressive coping and somatic diseases, Health Psychology 31(5), 2012.
- Soto, Perez, Kim, Lee and Minnick, Is expressive suppression always associated with poorer psychological functioning? A cross-cultural comparison between European Americans and Hong Kong Chinese, Emotion 11(6), 2011.
- Butler, Lee and Gross, Emotion regulation and culture: Are the social consequences of emotion suppression culture-specific?, Emotion 7(1), 2007.
- Le and Impett, When Holding Back Helps, Psychological Science 24(9), 2013.
- Min, Clinical Correlates of Hwa-Byung and a Proposal for a New Anger Disorder, Psychiatry Investigation 5(3), 2008.
- Suh and colleagues, How Suppressed Anger Can Become an Illness: A Qualitative Systematic Review of the Experiences and Perspectives of Hwabyung Patients in Korea, Frontiers in Psychiatry 12, 2021.
- Köllner and Schultheiss, Meta-analytic evidence of low convergence between implicit and explicit measures of the needs for achievement, affiliation, and power, Frontiers in Psychology 5, 2014.
- Kunda, The case for motivated reasoning, Psychological Bulletin 108(3), 1990.
- Venkit, Li, Pruksachatkun and Wu, The Need for a Socially-Grounded Persona Framework for User Simulation, arXiv 2601.07110, 2026.
- Zhang and colleagues, Context-Value-Action Architecture for Value-Driven Large Language Model Agents, arXiv 2604.05939, 2026.
- Bhattacharyya and colleagues, How Well Do Large Language Models Capture Human Personality?, arXiv 2606.18263, version 1, 2026.
- Yeo and Jaidka, Beyond Context to Cognitive Appraisal: Emotion Reasoning as a Theory of Mind Benchmark for Large Language Models, arXiv 2506.00334, 2025.
- Countless Lab, framework layout test on one designer’s 56 later projects, code in
research/framework-structure, and selector learning curve on Jester joke ratings, code inresearch/selector-curve.