Gender-Marking -ess: The Suffix that Failed

Abstract

Abstract Languages with grammatical gender typically have affixes, sometimes called motion affixes, for converting grammatically masculine, semantically male human nouns into grammatically feminine, semantically female ones (e.g. German -in, as in Professor “male professor”, Professorin “female professor”). It it widely assumed that the English suffix -ess (as in actress, waitress) is such an affix. We show that this assumption is false. By the time the suffix entered the language via French and Latin loanwords from the 13th century onwards, English had lost its grammatical gender system and human nouns without gender marking were no longer interpreted as male by default. While the suffix initially served a function in deriving female correlates to lexically male human nouns (for example, nouns referring to members of the clergy or the so-called nobility), it gradually shifted to a suffix invoking female-directed stereotypes. This is a fundamental difference to motion affixes and ultimately led to the disappearance of -ess as a productive derivational morpheme. Based on an empirical analysis of roughly 780 derivations and loanwords documented in the OED and various large corpora, we trace this shift and discuss it in terms of morphological theory, in particular, with respect to the notion of productivity.