Matches and mismatches in nominal morphology and agreement: Learning from the acquisition of Eegimaa

Funded by the ESRC and AHRC (ES/P000304/1)
April 2017-March 2020


Project overview

Theoretical accounts of how children learn the structures of words and grammatical features of languages differ considerably, but our knowledge of what is possible is limited by the existing focus on a relatively small number of languages associated with industrialised nations. This project investigates the acquisition of inflectional morphology, i.e., grammatical features and structures as reflected in word forms and associated grammatical agreement, in Gújjolaay Eegimaa, a language of the Atlantic family of the Niger Congo phylum spoken in Southern Senegal. Gújjolaay Eegimaa has a gender system of the type traditionally known as a noun class system. Noun class systems with complex gender agreement are characteristic of the Niger-Congo languages.

In Eegimaa nouns use prefixes to form singular and plural. For example ba- is the singular marker for ba-ginh 'chest', but its plural marker is u- as in u-ginh 'chests'. Nouns which have the same singular prefix, e.g. ba-, can form their plural with a different marker (e.g., bá-jur 'young woman', plural sú-jur 'young women'). Eegimaa has a complex morphological system of gender and number marking which is also reflected in its agreement system. Current knowledge as to how children acquire gender/noun class marking and agreement is based entirely on the Bantu languages of the Niger Congo family. There are no studies available of Atlantic languages which, though similar to Bantu in some ways, also have important differences.

Our project aims to answer questions about the strategies children rely on to learn complex language structure, the role of adult input language in the acquisition of morphology, how children cope with variation in language input from their caregivers, and the order in which they learn different noun class markers. The research involves both a longitudinal study (involving five children aged two to four over three years) and a cross-sectional study (ten each at ages three and four years). This project will thus be able to provide both an in-depth look and a broader overview of the acquisition process in this under-investigated area.

Children who participate in this project come from four of the ten villages of the Eegimaa speaking area. The longitudinal study involves recording children every 15 days for 30 minutes on video, with caregivers, older siblings, members of their extended families or other members of the community who interact regularly with them. Local assistants have been trained to record and transcribe data using ELAN. As of 22 October 2018, at least 60 hours of recording are being transcribed and coded for analysis. Preliminary results from the very first hours of recording have been presented at two conferences; the Child Language Symposium (CLS) 2018 and The World Congress of African Linguistics (WOCAL)2018, and in a departmental colloquium at the University of Essex and at the Workshop on Computational Approaches to Morphologically Rich Languages, University of Leeds in the Summer 2018.


Project outputs can be found here.