Differential object indexing in Maltese – a corpus based pilot study

Abstract

This paper presents the first corpus-based study of DOI in Maltese. In this pilot study, the potential triggering factors were tested as predictors in a descriptive model. The results show that the strongest predictor for object indexing in Maltese is word order, but when taking only semantic referential features into account, the analyses reveal that DOI seems to be strongly predictable by definiteness, as well as by the part of speech of the head of the NP. Our study therefore supports observations from previous investigations, both on Maltese and typological; furthermore, the analysis gives insight into the combined effects of the relevant factors.

Lexical anaphora : a corpus-based typological study of referential choice

“…3 As noted earlier, a substantial proportion of work in corpus linguistics is based exclusively on written data; for investigations into discourse structure and referential choices such as this one, the use of natural, unscripted spoken (or signed) data is essential, however, as only here are "considerations of language processing and efficiency-related trade-offs playing out under relevant time constraints of production and comprehension as well as noisy channel and similar considerations of the articulation bottleneck, principles of inference, and so forth" (Schnell et al 2021a: 10). For instance, Just & Čéplö's (2022) study of bound object indexing in Maltese had to switch from using written Universal Dependency corpora to using spoken texts as the feature under investigation, while widespread in speech, proved too rare in writing for rigorous analysis. In a similar vein, Schnell & Schiborr (2022: 178-179) note that the kind of long, complex NPs that drive the initial claims on dependency length minization and its effects on word order regularities in Futrell et al (2020b), while reasonably common in written texts, are barely present in spoken data; the greater tolerance for long NPs in written texts likely results from greater ease of written compared to spoken text comprehension, as in the former readers can adjust their reading speed at will, reread passages as needed, and so on, which not possible in spoken language.…”

Variable index placement in Gutob from a typological perspective

“…This term is misleading and suggests randomness in the choice of one marking strategy over another. However, usage based studies have shown that even though certain constructions might not be put down to a single factor which is easy to discern, their use is not random, but the respective form serves an intentional communicative goal on the part of the speaker (see Schikowski 2013 on DOM in Nepali, Just & Witzlack-Makarevich (accepted) on differential P indexing in Ruuli, or Just & Čéplö (2022) on the same phenomenon in Maltese). But even when acknowledging that a particular marking strategy is not unpredictable but employed for discourse-structuring purposes, one should be aware of the fact that the use of information-structural categories like topic and focus might bar the way to fair descriptions of the respective language system.…”