The project on South Slavic will be lead by Prof. Christina Manouilidou, based at the Uv Ljubljana, and will focus on a comparison between Slovenian and Bosnian/Croatian/Serbian (BCS). Manouilidou and her colleagues (Manouilidou et al 2016) have conducted behavioural research using the morphosyntactic vs. morphosemantic violation paradigm that is at the heart of the SaVANT project, and replicated the category vs. argument structure rejection rate and response time patterns we previously observed in Greek and English (Manouilidou & Stockall 2014).
In the 2016 study, Manouilidou et al investigated Slovenian agentive -ec deverbal nominals (eg. plavalec ‘swimmer’). BCS has a very similar morpheme, -ac, which has roughly the same restrictions on its distribution (it requires an agentive verbal stem). In additional to -ec/-ac, we also propose to investigate the Slovenian and BCS suffixes -telj/-telj. The -ec/-ac suffixes attach to derived ‘l-participles‘ in both languages, while the -telj suffixes attach to verbal stems. This difference in the size and complexity of the stem may be associated with differences in the speed and accuracy of the processing responses.
In addition to this grammatical difference in the distribution of the two suffixes, which is consistent across Slovenian and BCS, there are differences in the frequencies of the suffixes within and across languages. In Slovenian, -ec formations are considerably more frequent (cf. Bajec et al. 2014); l-participle-based -ec formations are the most productive agentive deverbal nominalizations (Stramljič Breznik 2010). By contrast, in recent vocabulary, -telj formations have been claimed to be primarily restricted to one out of four or more verb classes (e.g. Voršič 2013). In BCS, -telj appears to be more productive and frequent than its Slovenian counterpart (cf. Babić 2002 vs. Bajec et al. 2014, Stramljič Breznik 2010). Thus comparison between and within these two languages also allows us to investigate the role of frequency and productivity in morphosyntactic and morphosemantic processing.