The Theory of Constrained Identity: Identity Conditions Upstream, Typicality Downstream
When emotion science gave up the search for emotional essences, it also lost an answer to a prior question: what makes an episode of anger an episode of anger rather than a useful guess under a familiar name? I argue that several influential accounts, among them the theory of constructed emotion, relocate identity conditions downstream into similarity, prediction, categorization, and agreement, and that this relocation leaves error, disagreement, and revision without a standard. I propose the Theory of Constrained Identity (TCI) as the repair.
It orders four layers with a direction of dependence: a constraint layer that fixes identity conditions and admissibility, an instance layer that realizes them under determinate values, a category layer that groups admitted instances by similarity and typicality, and a goal layer that evaluates both instances and constraint structures without supplying identity. Five invariants keep the layers apart. The arrangement dissolves the classical–prototype dichotomy by assigning each theory its level.
I apply the TCI to the folk concept of anger in contemporary American English and use it to separate misapplication, disagreement, cultural variation, and revision from drift. What is new is a strict four-layer order, a separate goal layer, and a test that makes the order inspectable in executable code. Working paper 1 of 8 in the series on the frameworks of How to Design an Emotion: The Deconstruction and Computation of Emotional Meaning from First Principles (Universal Moral Concepts III, New Intellectual LLC, 2026; ISBN 979-8-9898135-9-9 paperback, 979-8-9898135-3-7 Kindle).
The book is the full treatment. Book page: https://newintellectual.com/how-to-design-an-emotion ; related: Part II . Buy on Amazon .
