engleski

Conjunct and single NP subjects: a comparison of naïve and expert intuitions

The results of an experimental picture matching study (Willer-Gold et al, in preparation) lead to the conclusion that closest conjunct agreement (CCA) in South Slavic languages is not a result of ellipsis of a biclausal structure as no major differences were noted in ratings for conjunct subjects and single NP subjects, in line with previous studies on CCA in South Slavic languages (Marušič et al. 2015 ; Willer-Gold et al. 2016). The original experiment (Willer-Gold et al, in preparation) was conducted among a group of undergraduate students, i.e. naïve participants (in the sense of Dąbrowska 2008, 2010). In the present study, the same experiment is conducted with a group of expert par-ticipants of comparable size and dialectal background, Croatian language teachers and linguists (persons involved at the primary, secondary and tertiary level of education). In this study, the same stimuli have been used as in the original experiment, thus contribut-ing to the discussion on the methodology of collecting empirical linguistic data (inter alia, Spencer 1973 ; Ferreira 2005 ; Culbertson, Gross 2009 ; Devitt 2010 ; Phillips 2010 ; Gross, Culbertson 2011 ; Gibson, Fedorenko 2013 ; Sprouse, Schütze, Almeida 2013).Studies conducted by Dąbrowska (2008 with non-linguists as participants and 2010 with linguists as participants) showed that judgements of the same sentences by lin- guists differ systematically from those provided by non-linguists. The study presented in our paper is in line with Dabrowska’s 2010 study, replicating the experimental study by Willer-Gold et al. (in preparation). One of the insights of Dąbrowska’s 2010 study was that the difference in acceptability ratings between naïve and expert participants could be explained by the amount of exposure to sentences with long distance depend- encies between the two groups of participants. This paper discusses whether a similar correlation between the level of expertise and the participants’ linguistic intuitions for both conjunct and single NP subjects can be found. While we do not assume that the dif-ference between the two experimental groups can be explained by satiation effects (cf. Snyder 2000), the data collected in the present study indicates a difference in speaker 120intuitions between the two groups, noticeable primarily in terms of lower average ac-ceptability ratings, both for critical conditions and fillers, and in terms of higher degree of variability in acceptability ratings for all items.

Closest Conjunct Agreement, Croatian, naïve vs. expert participants, native speaker intuitions, sentence-picture matching experiment

nije evidentirano