When Nouns Turn Negative: The Evaluative Absorption Model

Pejoration describes semantic change whereby words shift from neutral or positive to negative over time. We propose the Evaluative Absorption Model, both to characterize a specific pejoration phenomenon and to operationalize it by tracking (1) an initial rise in explicit negative modifiers in attributive position as a word acquires a negative sense, and (2) a subsequent decline in such modifiers once that sense becomes conventionalized. We illustrate this trajectory for four nouns — “propaganda”, “manipulation”, “harassment”, and “discrimination” — using diachronic corpus data from Google Ngram. Our model offers a transparent method for quantifying semantic pejoration.