Skip to main content

Emotion Taxonomy from ∇P\nabla P

An emotion label is an observation when a participant supplies it under a stated question. Heart rate, approach or avoidance, and choice under uncertainty are other observations. Before proposing that any combination identifies an emotion, specify the person, time interval, alternatives and prediction rule.

What purity change measures​

For a differentiable density matrix, P=Tr⁡Γ2P=\operatorname{Tr}\Gamma^2 gives P˙=2Tr⁡(ΓΓ˙)\dot P=2\operatorname{Tr}(\Gamma\dot\Gamma). A Hamiltonian commutator contributes zero by cyclicity. In a reset flow Γ˙=k(σ−Γ)\dot\Gamma=k(\sigma-\Gamma) the sign depends on Tr⁡(Γσ)−P\operatorname{Tr}(\Gamma\sigma)-P. Resetting toward a pure state, starting from a partly mixed state on the line joining it to I/dI/d, can increase purity. Resetting toward I/dI/d decreases the purity of every nonuniform state. The same two effects occur in a prepared register with no emotion report. A sign assignment such as “increasing purity means pleasure” requires a separately measured outcome.

Testing a proposed correspondence​

Collect the participant’s ratings of valence and arousal separately from named emotion reports. Preserve the situational prompt, timing, response options and missing trials. To test a physiological predictor of fear, compare it with a baseline for high arousal and a baseline for features of the prompt alone. Use a new situation or participant held out from fitting. A match on arousal with no improvement in predicting the named report identifies the narrower quantity.

Reporting and interpretation​

Confidence in naming an emotion can be assessed independently of agreement with another observer. If a model assigns a density matrix, its encoder and phase references must be fixed before testing whether state changes predict either outcome. Similarity of reports across people supports a common rating structure within that protocol. Identifying their felt qualities requires additional evidence. Measurement model · Phenomenal distinctions