About

Human emotional vocabulary averages 30 - 40 distinct labels. Research by Cowen & Keltner (2017) identified at least 27 distinct emotional categories, while Plutchik's wheel models 8 primary emotions with combinatorial derivatives. Most people default to 5 - 7 habitual descriptors, creating a compression artifact in self-awareness. This generator draws from a dataset of over 120 catalogued emotions spanning 6 taxonomic categories: Primary, Secondary, Complex, Social, Cognitive, and Existential. Each entry includes intensity calibration on a 1 - 5 scale, opposing emotion pairing, and a brief psychological note grounded in affective science.

Applications include creative writing prompts, emotional literacy training, therapy icebreakers, acting exercises, and journaling. The tool approximates Plutchik's dimensional model but does not replace clinical assessment instruments such as PANAS or the Geneva Emotion Wheel. Note: emotion categorization remains debated in affective neuroscience. Classifications here follow consensus from Ekman, Plutchik, and the componential appraisal tradition.

Formulas

Emotion selection uses cryptographically strong randomness via the Web Crypto API, filtered by user-selected category constraints. The selection probability for any emotion e within filtered set S is uniformly distributed:

P(e) = 1|S| for all e ∈ S

where |S| is the cardinality of the filtered emotion set. When all categories are active, |S| = 120. The random index i is computed as:

i = floor(r × |S|)

where r ∈ [0, 1) is derived from crypto.getRandomValues. The slot-machine animation uses cubic deceleration easing:

v(t) = v₀ × (1 − t)³

where v₀ is initial scroll velocity, t ∈ [0, 1] is normalized time, yielding smooth deceleration to the target emotion.

Reference Data

Emotion	Category	Intensity	Opposite	Valence
Joy	Primary	4	Sadness	Positive
Fear	Primary	5	Anger	Negative
Anger	Primary	5	Fear	Negative
Sadness	Primary	4	Joy	Negative
Disgust	Primary	4	Trust	Negative
Surprise	Primary	3	Anticipation	Neutral
Trust	Primary	3	Disgust	Positive
Anticipation	Primary	3	Surprise	Positive
Nostalgia	Complex	3	Indifference	Mixed
Schadenfreude	Social	2	Compersion	Mixed
Ennui	Existential	2	Enthusiasm	Negative
Euphoria	Secondary	5	Despair	Positive
Melancholy	Complex	3	Elation	Negative
Awe	Complex	4	Contempt	Positive
Gratitude	Social	3	Resentment	Positive
Envy	Social	4	Admiration	Negative
Shame	Social	5	Pride	Negative
Pride	Social	3	Shame	Positive
Curiosity	Cognitive	2	Apathy	Positive
Confusion	Cognitive	2	Clarity	Neutral
Serenity	Secondary	1	Anxiety	Positive
Anxiety	Secondary	4	Serenity	Negative
Contempt	Social	3	Awe	Negative
Dread	Existential	5	Hope	Negative
Hope	Existential	3	Dread	Positive
Saudade	Existential	4	Contentment	Mixed
Epiphany	Cognitive	4	Bewilderment	Positive
Ambivalence	Cognitive	2	Conviction	Neutral
Guilt	Social	4	Innocence	Negative
Compassion	Social	3	Cruelty	Positive

Frequently Asked Questions

The dataset contains over 120 emotions organized into 6 categories: Primary (Ekman's basic emotions), Secondary (blends and gradations), Complex (multi-component states like nostalgia), Social (interpersonal emotions like shame and gratitude), Cognitive (epistemic states like curiosity), and Existential (meaning-related states like dread and saudade). The taxonomy draws from Plutchik's wheel, Ekman's basic emotion theory, and the componential appraisal framework.

Intensity reflects the typical arousal level associated with the emotion in affective science literature. A value of 1 (e.g., Serenity) indicates low physiological arousal and mild subjective experience. A value of 5 (e.g., Fear, Shame) indicates high arousal with strong autonomic nervous system activation - elevated heart rate, cortisol release, and behavioral urgency. These values are approximate consensus figures, not clinical measurements.

Emotions like Nostalgia, Schadenfreude, and Saudade involve simultaneous positive and negative appraisal components. Nostalgia combines pleasant memory recall with the pain of loss. Schadenfreude pairs pleasure with awareness of another's misfortune. The valence field reflects the dominant research classification from Barrett and Russell's circumplex model, where mixed-valence states occupy regions between the positive and negative poles.

By default, consecutive duplicates are prevented. The algorithm checks the last generated emotion and rerolls if a match occurs, up to 3 attempts. If the filtered set contains only 1 emotion, duplication is unavoidable. The history panel tracks up to 50 prior results to help identify patterns.

The generator uses crypto.getRandomValues() from the Web Crypto API, which provides cryptographically secure pseudorandom numbers sourced from the operating system's entropy pool. Math.random() uses a deterministic PRNG (typically xorshift128+) that is predictable if seeded. For emotion generation the practical difference is negligible, but crypto randomness eliminates any sequential correlation artifacts that could bias repeated draws.

Opposite pairings follow Plutchik's dyadic model where emotions are arranged on a wheel with polar opposites: Joy-Sadness, Trust-Disgust, Fear-Anger, Surprise-Anticipation. For Complex, Social, and Cognitive emotions outside Plutchik's original 8, opposites are assigned based on appraisal theory - matching emotions with inverted appraisal dimensions (e.g., high control vs. low control, self-caused vs. other-caused). Some pairings are approximate where no consensus opposite exists.