I propose a novel analysis of differential object marking (DOM) that revolves around nominal licensing and, unlike most previous accounts, does not appeal to object visibility (e.g., Danon, Lyutikova & Pereltsvaig), object raising (e.g., Bhatt & Anagnostopoulou, Baker & Vinokurova), or object differentiation or identification (e.g., Aissen, De Hoop & Malchukov). I argue (i) that not all nominals need abstract licensing (following Massam, Danon, Ormazabal & Romero), but also that all nominals have the potential to be licensed, that is, all nominals are visible to case and agreement processes; (ii) that the set of objects that are overtly marked (e.g., by case or agreement) can reveal the set of nominals that require licensing in a language, and (iii) that clauses typically have one obligatory nominal licenser, with secondary licensers merging only when needed for convergence (following Levin & Massam; Bobaljik; Laka; Rezac). Taken together, I show that variation in the types of nominals that need licensing and the location/identity of obligatory and secondary licensers can derive crosslinguistic differences in case, agreement, and DOM patterns. This unified account simplifies our understanding of nominal licensing within and across languages, as it does not require objects to have special properties as compared to subjects, nor does it fundamentally differentiate DOM languages from non-DOM languages. The motivating data come largely from the Neo-Aramaic language Senaya, which clearly illustrates that certain nominals can occupy a position where abstract licensing is unavailable.
All Science Journal Classification (ASJC) codes
- Language and Linguistics
- Linguistics and Language