On the limits of etymology

ABSTRACT Using the classical example of -ize versus -ise in English as a case study, this article argues that insight into etymology, contrary to an assumption implicit in some dictionaries, cannot be of much help in guiding spelling, nor can arguments concerning spelling be meaningfully substantiated on the basis of knowledge of etymology. In building this argument, I compare the original Greek senses of -ίζω -izɔ· to the usage of this suffix when borrowed into Latin, showing how Latin language users have made creative use of elements taken from Greek, integrating them into the language-specific structure of Latin. English speakers have reinterpreted and integrated the suffix -ize/-ise in language usage and structure in similar creative ways by drawing on Greek, Latin and French, meaning that a modern English verb spelled with -ize or -ise can neither be identified as ‘Greek’, ‘Latin’ or ‘French’ by the ordinary language user. Hence, a reference to a word’s origin is not a safe guideline for deciding how it should be spelled.