Color is shared. Meaning has an accent.

Across 132 studies, 42,266 people, and 64 countries, humans keep finding a shared color grammar. Language, climate, context, and culture give it an accent.

Ask what red means to someone and the honest answer is not one emotion. Red is anger, love, danger, desire, power, heat, and celebration. Ask what sadness looks like and the answer is not one color either. It is black, grey, or a dark, desaturated blue, depending on the task and the culture.

That many-to-many structure is the central result of the color-emotion literature. The associations are too consistent to be arbitrary and too flexible to be a universal dictionary.

Our review follows this structure across the fields of psychology, affective science, psycholinguistics, perception, environmental research, and clinical work. Color has emotional meaning, yes, but are these meanings universal across cultures? Where do these meanings come from, and why?

The modern synthesis comes from Jonauskaite and Mohr’s review of 128 years of research, where they gathered 132 peer-reviewed studies covering 42,266 participants across 64 countries. Across the studies, a wide array of methodologies were used, and the stimuli presented to participants also varied. Some showed calibrated color patches, while others showed color words, probed discrete emotions like anger or joy, or asked about things like valence, arousal, power, preference, or mood.

Even with the different methodologies, orderly aggregate patterns emerged. Studies found that in addition to hue, factors like lightness/darkness, saturation, warmth/coolness, and even surrounding context mattered as well for emotion association. For example, light colors were associated with positive affect, dark colors with negative affect, saturated colors with greater arousal and power, and desaturated colors with subdued or negative states. General color families also shared similar meanings, like yellow and orange clustering around positive, high-arousal meanings, while blues and greens were positive but low-arousal. Greys were negative and low-arousal, while black mapped to negative affect and, in some contexts, high power or threat as well. Depending on context, reds could be either positive or negative valence, while remaining consistently activating and forceful.

So we learn that two reds can behave differently depending on context, saturation, warmth, or lightness, meaning that a bright, saturated red can feel energetic or celebratory, while a dark red may feel heavy or controlled. Similarly, pale blue may be calm and relieved, but dark, desaturated blue can be more associated with sadness. The color name can also stay constant while its affective coordinates move. The rest of this post will dig deeper into the specific nuances of these associations, as well as the factors that tend to affect them.

Lightness, saturation, and temperature do much of the work

Perhaps not surprisingly, the dimensions that color research converges on are similar to the dimensions that organize emotion research too. For example, emotion research often incorporates valence, arousal, and dominance. Valence asks whether an experience is positive or negative, arousal asks whether it is activated or subdued, and dominance, sometimes called power or control, asks whether the person feels agentic and in control or overpowered. Color research shows parallel dimensions, where lightness shifts evaluation (valence), saturation reflects intensity (arousal), and warmth or redness often corresponds to activation and power (a proxy for dominance).

Gao and colleagues tested 214 color samples across seven regional groups and found that a color’s chroma and lightness explained more of the affective ratings than hue or cultural background. Take red, for example. Red does not reliably separate positive from negative emotion, because love and anger can both be red, but it does separate intensity and activation from quiet and calm. Red evokes blood, flushing, heat, warning, dominance, and action, so while different studies may disagree on the valence of red, they all find it to be associated with high intensity or arousal.

On the other hand, you might take the English phrase “feeling blue” to mean that blue means sadness across the board, but that is not the case. Sadness is not actually uniformly blue across languages, images, or even experiments. Instead, blue is seen as calm or comforting in many cultures, depending on saturation. A dark blue, however, which shifts the valence axis through color lightness/darkness, is found to be more associated with sadness.

Crucially, several distinctions matter more than the exact hue in associations with emotion. For example, anger is often more saturated than frustration, while sadness is often darker. Hope and relief may share a certain lightness but differ in warmth. These are just a few examples, but many findings have paralleled the general rule of thumb of lightness, saturation, and warmth corresponding respectively with valence, arousal, and dominance.

Jonauskaite and colleagues’ 2020 paper was the largest cross-cultural study on these patterns, testing 4,598 participants in 30 countries, with native speakers of 22 languages. Each participant connected 12 color terms with 20 emotion concepts. The average pattern similarity across nations was surprisingly high, with pink linked to love and pleasure, yellow linked to joy and amusement, and red linked to both love and anger (as mentioned above).

However, there were still differences among countries. Jonauskaite and colleagues’ study used a machine learning algorithm to find that nationality still predicted color-emotion associations after the shared pattern was accounted for. Additionally, countries were more similar when they were geographically closer to each other, or when their languages were more closely related. In other words, color seems to have a common structure that travels across locations, but local cultures and experiences change its accent.

Earlier work corroborates this general structure, with Adams and Osgood analyzing color meanings within a 23-culture semantic-differential project. Black and grey, dark colors, were broadly evaluated negatively, while red was affectively strong and active. Later, Hupka and colleagues compared anger, envy, fear, and jealousy in Germany, Mexico, Poland, Russia, and the United States. Black and red were robustly associated with anger, and black also with fear. Envy and jealousy were less stable, with German participants connecting them with yellow, while American participants connected envy with black, green, and red (think of the phrase “green with envy” in English), and Russian participants emphasizing black, purple, and yellow. Here, we can see how the general structure of emotion-color associations changes accents in different locales, with specific metaphors or lexical conventions perhaps influencing how a certain color or emotion is recognized within a certain culture.

Like with most things, the physical environment of any given phenomenon also matters. In a 55-country study from Jonauskaite and colleagues in 2019, the association between color and climate was studied across 6,625 participants. The yellow-joy link, long treated as one of the settled ones, turned out to vary with where people live. Participants further from the equator and in rainier countries were more likely to call yellow joyful, and the likelihood ranged from 5.66% in Egypt to 87.68% in Finland.

The tempting explanation is sunshine, and the paper tests it directly. On its own, sunshine predicts the association at β = −0.435. Once distance from the equator and rainfall enter the model, sunshine falls to −0.119 and is no longer significant, while latitude nearly doubles. What is often summarized as a hot-and-dry effect is, as measured, a latitude-and-rainfall effect.

The result is not a simple latitude rule, though. Warm, humid countries do not always resemble warm, dry countries. What matters is the local ecology of the color, and how it gets contextualized in the conditions of a given place. This fits Palmer and Schloss’s ecological-valence theory, where people come to like or dislike colors partly through their average emotional response to the things carrying those colors in their everyday environment. Clear water and open sky support positive blue, while rot and contaminated waste can make brown or yellow-green have negative connotations. Things that are universal across bodies and environments, like blood being red or darkness obscuring vision, also tend to have more universal color-emotion associations. More specific to a locale’s social environment and institutions, things like school colors, national colors, religious colors, and political colors can also shift or modulate these associations.

Delving deeper into how language affects this color-emotion grammar, psycholinguists have found that a person’s language background can also affect their associations. Color terms activate idioms, metaphors, objects, and social knowledge. English speakers “see red”, “feel blue”, and (as mentioned above) “turn green with envy”, while the French and German languages repeat some of these mappings but replace others. Lithuanian descriptions, on the other hand, can grade anger through a sequence of colors, moving from white to red to blue to black as intensity rises. These all matter, to some degree, in building the associations people have between emotion and color.

For example, Soriano and Valenzuela’s work on Spanish color terms showed that emotional associations can appear in implicit lexical processing, not only in explicit questionnaires. Additionally, Hupka’s cross-cultural results similarly suggest that mythology and literature of a certain culture can also contribute to the mappings participants report.

However, we should be careful about how strong we make our claims here. These findings are not to say that linguistic determinism is at work, but rather that there are some effects, to some degree, that we see from language background, just as we do from things like physical and social environment.

Additionally, liking a color, labeling an emotion (for example, in a psycholinguistic task), and feeling an emotion are distinct processes. But three claims are often collapsed:

1. People associate a color with an emotion.
2. People prefer a color.
3. Exposure to a color changes their emotional state.

If we actually look into this, the evidence is strongest for the first claim. Evidence is substantial but context-sensitive for the second, and it is weaker and more conditional for the third.

Elliot and Maier’s 2014 review of color psychology argues that color can influence cognition but repeatedly emphasizes context and boundary conditions. For example, red can signal error and avoidance in an achievement task, attraction in an affiliation context, or danger in a warning system, meaning that the physical stimulus can remain the same but the meaning can change if the task changes.

The 128-year review reaches a similar limit. Most studies measure abstract correspondence, where participants selected an emotion for a color or rated its valence, arousal, or power. While these are associations, those responses do not establish that the color produced the reported emotion. Researchers warn against deterministic relationships between color and emotion, and advise that we take these findings, and their implications, in the context of the task and environment.

While we don’t have one definitive theory for why emotions acquire certain colors, we can combine accounts from studies of perception, the body, the environment, and language and culture to get a holistic picture of this phenomenon.

1. Perceptual structure. Lightness and saturation carry broad affective information before hue does much of the specific work. Specifically, light colors tend to feel more positive or open, saturated colors more activated, and dark or desaturated colors quieter and heavier. This helps explain why neighboring emotions can share a hue but differ in intensity.

2. Ecological learning. People learn colors through the things that bear them. Colors of nature inherit value from the corresponding things in nature, like lush green vegetation. These environmental regularities are widely shared, but their value changes with local conditions, such as the yellow example mentioned above.

3. Embodied signals. Emotion changes the visible body too. Anger, exertion, attraction, and embarrassment can increase facial redness, while fear, illness, and fatigue can produce pallor. Because observers use facial coloration to read emotion, red may acquire meanings of activation, heat, urgency, and social exposure without being limited to any one emotion. This is the weakest of the four accounts, and the underlying evidence is thin: the study showing that people can identify emotions from facial color alone ran its human-judgment component on 20 participants, and the follow-up asking how much red actually helps anger recognition answered its own title by questioning the practical significance of the effect.

4. Language and metaphor. Expressions such as a dark mood, a bright future, seeing red, or feeling blue repeatedly connect visual properties with emotional states. Some may extend bodily or environmental experience, while others are simply conventions that languages teach and reinforce.

Together, these mechanisms produce families rather than fixed translations. On top of the shared grammar of lightness, saturation, and warmth, hue supplies a culturally accented vocabulary that is modulated by cultural and physical context.