Birzer, SandraSandraBirzer0000-0003-4167-9825Mayer, DanielaDanielaMayer2021-09-272021-09-272021https://fis.uni-bamberg.de/handle/uniba/50630We conducted an online experiment with speakers of Russian as primary and secondary language in order to identify the factors which influence co-reference choice for AP clauses. The experiment aimed to test three linguistic (i.e. syntactic and semantic) and five sociolinguistic hypotheses: 1) the co-reference is assigned to the matrix verb argument that is closer to the AP in terms of syntactic linearity; 2) due to their special semantic structure, the assignment of co-reference to the second matrix verb argument is more likely for object control verbs than for verbs with “plain” semantics or for subject control verbs; 3) semantic neutrality of the matrix sentence fosters ambiguous co-reference. 4) RSL speakers have longer reaction times than RPL speakers, because it is more difficult and thus time-consuming for them to analyze ambiguous co-references. 5) RSL speakers are more prone to use the syntactic position of the AP as criterion for their co-reference choices than RPL speakers. 6) Respondents with a high level of education are more likely to establish co-reference with the first argument of the matrix verb. 7) Respondents older than 40 years are more likely to establish co-reference with the first argument of the matrix verb. 8) Women are more likely to establish co-reference with the first argument of the matrix verb. For data analysis we used a logistic regression model based on the data of 67 RPL speakers and 45 RSL speakers. The three linguistic hypotheses (H1-H3) have been verified for both speaker groups. We may conclude that the linear closeness between the AP and the second argument of the matrix verb – prototypically realized as sentence-final position of the AP – opens up the possibility to establish co-reference between the two. However, the possibility is made use of only if the sentence semantics is “neutral”. Additionally, an object control verb in the “neutral” setting increases the likelihood of co-reference with the second argument of the matrix verb, so the ideal scenario for this co-reference choice is a sentence-final AP, accompanied by an object control verb as matrix verb and “neutral” sentence semantics. Thus, the syntactic position of the AP is the prerequisite, but sentence and verb semantics are the decisive factor for the co-reference choice. For our two speaker groups we observed a statistically non-significant tendency of RSL speakers to choose the second argument of the matrix verb as co-reference controller more often than RPL speakers do. [...]engadverbial participle, Russian, primary language, secondary language, ambiguity, co-reference, experiment490preprinturn:nbn:de:bvb:473-irb-506307