Optimal categorisation: the cognition of nominal classification systems

Objectives: Nominal classification systems (gender, classifiers) are a functional means of categorisation that vary enormously across languages. Closely related Oceanic languages of Melanesia show staggering variation in their number and type of classifiers. How does the Iaai language carve up nouns into 23 semantic groups whilst the Merei language uses only two? What implications do these vastly different systems have for the cognitive representations of their related concepts? This ESRC-funded project exploits this variety to reveal the relative cognitive efficiency of gender and classifier systems, as a route to understanding optimal categorisation.

Design: We combine typological enquiry and psycholinguistic experimentation across seven complementary experiments (free listing, card sorting, video vignettes, possessive labelling, eye tracking, storyboards, category training), which compare nominal classification systems in six Oceanic languages of Vanuatu and New Caledonia.
Methods: During the first field trip to Vanuatu and New Caledonia, 132 participants (22 native speakers from each of the six languages) take part in free listing, card sorting and video vignette experiments.

Results: We present initial data from the first three experiments. Free listing establishes central members of a classifier’s semantic domains, which vary across languages. Card sorting reveals how speakers categorise relevant entities and whether conceptual groupings map onto classifiers. Finally video vignettes ascertain whether there is rigid or flexible assignment of nouns to classifiers.
Conclusions: We discuss how these experiments uncover the nature of nominal classification systems, comparing objective data across languages and experimental contexts to reveal a model for optimal categorisation.