Project: Variation in gender agreement in modern Swedish
The aim of this project is to find out if Swedish speakers use one or two systems for gender agreement in language processing. There are currently two theoretical approaches: one claims that there is only one system for agreement and that system is semantic, the other claims that there are two systems, one semantic and one syntactic. We try to find out which of these approaches is the correct one.
Project manager Fredrik Heinat, Linnaeus University Other project members Eva Klingvall, Lund University Participating organizations Linnaeus University and Lund University Financier Erik Wellanders Fond Timetable 1 July 2020 – 30 June 2021 Subject Psycholinguistics, Linguistics, English, Swedish (Faculty of Arts and Humanities, Department of Languages)
More about the project
There is no doubt that agreement is governed by syntactic factors, e.g. a head determines the forms of its modifiers. But there are examples of semantics influencing the agreement system too.
According to Josefsson (2014) there are two approaches to the Swedish agreement system: 1) there is only one system and it is semantic in nature, and 2) there are two systems, one semantic and one syntactic. The aim of this project is to try to obtain data that can tell us which of these approaches is the most likely one from a language processing perspective. In Swedish, we find agreement predominately in the noun phrase. Regarding referential/semantic gender we find clear distinctions in definite NPs. In (1) we see on the definite article den 'the' that the lexical gender of the nouns is non-neuter, but in (1c) the adjective gammal 'old' optionally inflect for the male referent of the head noun.
(1a) den gaml-a kvinnan the old-def. woman
(1b) den gaml-a stolen the old-def chair
(1c) den gaml-e/gaml-a mannen the old-m.def./old-def. man
In order to investigate what Swedish speakers’ reaction to gender agreement are, we will conduct an estimation study. We investigate how changing the inflections on the adjectives influences native Swedish speakers' acceptability to the NPs. The findings from this study will be used in a later EEG-study where we investigate if the ratings are reflected in the real time processing of the NPs, and if that can tell us anything about what agreement systems are at work, only a semantic one, or a semantic and a syntactic.