Core affect
Also known as: Circumplex model of affect
The ever-present background feeling, mapped by pleasantness and energy.
What it means
Core affect is the most elementary, always-present neurophysiological feeling state, describable as a single point in a two-dimensional space defined by valence (pleasant to unpleasant) and arousal (activated to deactivated). On the circumplex model, named emotions are arranged around this circle, so excitement is high-arousal pleasant, contentment is low-arousal pleasant, and depression is low-arousal unpleasant. Core affect is free-floating and need not be about anything, becoming a full-blown emotion only when attributed to an object and elaborated by appraisal and context. The framework underlies dimensional theories that treat emotions as constructed from these basic ingredients rather than as discrete biological categories. It is widely used to measure mood and affect in research.
How it is measured
The circle is derived empirically, not assumed. Russell asked people to sort emotion words and scaled the similarity judgments, and multidimensional scaling repeatedly returns a roughly circular two-dimensional arrangement. To capture a person's momentary core affect, researchers use self-report grids: the Affect Grid asks for a single mark in a nine-by-nine square whose axes run pleasant to unpleasant and sleepy to aroused, or separate valence and arousal ratings collected through the nonverbal Self-Assessment Manikin. Because it needs only two numbers, core affect is cheap to sample many times a day, which is why experience-sampling and mood-tracking studies lean on it. The trade-off is resolution: two coordinates locate a feeling on the wheel but say nothing about its cause or content.
The rotation debate
Two traditions disagree about which axes are fundamental. Russell puts valence and arousal at the horizontal and vertical. Watson and Tellegen rotate the same space by roughly forty-five degrees and read it as positive activation and negative activation, the axes behind the widely used PANAS scales, where being enthusiastic and alert and being distressed and jittery are treated as separable systems rather than opposite ends of one pleasantness line. Statistically the two solutions describe the same cloud of data; they differ over which rotation is psychologically real, and whether pleasant and unpleasant feelings are strictly bipolar or can co-occur. The choice is not cosmetic: adopting valence-arousal versus positive-negative activation changes which items you administer and how you score mixed or blended emotions.
What the evidence shows
A clean picture of two orthogonal axes implies valence and arousal should be uncorrelated, and averaged across many people and moments that near-independence roughly holds. But it is not a law. Kuppens and colleagues modeled eight datasets and found that, on average, arousal traces a weak V-shape against valence: feelings grow more activated toward either the pleasant or the unpleasant extreme and are calmest in the neutral middle. More telling, the relation varied widely between individuals, strongly coupled for some, flat or reversed for others, and it tracked personality and culture. Core affect therefore has no single fixed geometry. The circle is an average over people, and each person's version of it is itself a measurable, meaningful difference.
Where it shows up
Reducing feeling to two continuous dimensions lets core affect travel well beyond mood questionnaires. In affective neuroscience, valence is linked loosely to mesolimbic reward circuitry and arousal to ascending systems through the reticular formation and amygdala, though no clean one-region-per-axis mapping survives scrutiny. In clinical work, the shared low-valence region helps explain why depression and anxiety so often co-occur and resist separation early on. Developmental studies find that young children sort emotions by good versus bad before they reliably use arousal, so the valence axis matures first. And affective-computing systems that label music, faces or text with an emotion typically output a valence-arousal pair, borrowing the circumplex as a compact, language-independent code.
Examples
Waking up vaguely 'up and energized' with no particular cause is core affect; it becomes joy only once you tie it to the good news in your inbox.
Two coffees in, you feel wired but pleasant. The same jittery energy becomes 'excited' at a party and 'nervous' in a waiting room, depending on what you pin it to.
The 3pm slump, low energy, mildly unpleasant, about nothing in particular, sits where sadness sits on the map, which is why a dull afternoon can feel like a bad life.
A streaming service tags each track with a valence-arousal pair, then assembles a calm-focus playlist by pulling low-arousal, mildly pleasant songs and filtering out the anxious high-arousal ones.
A depression-tracking app pings you twice a day for one tap on a pleasant-to-unpleasant, calm-to-tense grid; weeks of those dots trace your core affect without ever naming an emotion.
First described in James Russell (1980, 2003); valence-arousal circumplex.
Key references
- Kuppens, P., Tuerlinckx, F., Russell, J. A., & Barrett, L. F. (2013). The relation between valence and arousal in subjective experience. Psychological Bulletin, 139(4), 917-940. doi.org/10.1037/a0030811
- Posner, J., Russell, J. A., & Peterson, B. S. (2005). The circumplex model of affect: An integrative approach to affective neuroscience, cognitive development, and psychopathology. Development and Psychopathology, 17(3), 715-734. doi.org/10.1017/S0954579405050340
- Russell, J. A. (2003). Core affect and the psychological construction of emotion. Psychological Review, 110(1), 145-172. doi.org/10.1037/0033-295X.110.1.145
- Russell, J. A., & Barrett, L. F. (1999). Core affect, prototypical emotional episodes, and other things called emotion: Dissecting the elephant. Journal of Personality and Social Psychology, 76(5), 805-819. doi.org/10.1037/0022-3514.76.5.805
- Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161-1178. doi.org/10.1037/h0077714