Processing and production of affixes in Georgian and English: Testing a processing account of the suffixing preference

Abstract

Abstract The hypothesis that affixes following a stem are easier to process than ones preceding has not been tested in a straightforward manner in any language, as far as we know. Cutler, Hawkins & Gilligan (1985) and Hawkins & Cutler (1988) adduce some evidence that supports this hypothesis indirectly, but they do not conduct experiments to test it directly. They use this hypothesis to explain in part the suffixing preference. Some others, such as Asao (2015), continue to assume the correctness of the hypothesis. We do not aim to explain the suffixing preference at all but to test the hypothesis that affixes preceding the stem (informally, prefixes) disrupt the comprehension of a word more than affixes that follow (informally, suffixes) do. In this paper we test this hypothesis (henceforth the ‘Cutler--Hawkins hypothesis’) on Georgian, because it has a wide variety of prefixes and suffixes, and in a single experiment on English. In Georgian we test a prefix and a suffix that mark the person of the subject in a verb, a circumfix and a suffix that mark derivation in nouns, and a prefix and a suffix that form intransitive verbs (usually called ‘passives’ in Georgian). Across the set of experiments, we find little support for the Cutler--Hawkins hypothesis.