Old French subject pronouns (Spro) were omissible if postverbal (Foulet 1928), but not freely so (Vance 1997, Zimmermann 2014). This article addresses their partial omissibility in discourse-syntax terms, following work on partial null subject languages by Holmberg and Nikanne (2002) and Modesto (2008). An observational study of dialogic responses in 13th century prose romances is first reported, finding strong indications of covariation between the Topic/Focus status of an initial non-subject constituent and the expression/omission of post-verbal Spro. A quantitative investigation, in such texts, of preposed discourse-linked anaphoric constituents and preposed intensifiers, taken as diagnostic of Topichood and Focushood respectively, confirmed this analysis. We take null Spro to be available (i) when a null Topic operator targets left-peripheral TopicP, and (ii) with a left-peripheral Focused expression. When a discourse-linked non-subject constituent occupies TopicP, however, Spro must be overt.

This paper investigates the relationship between referential and non-referential null subjects in Middle Norwegian. It argues that overt, non-referential subject det arose before the loss of referential null subjects, contrary to the predictions of much previous work. A diachronic analysis compatible with the empirical findings is sketched out.

In this self-paced reading study, we first tested the cross-linguistic validity of the position of antecedent strategy proposed for anaphora resolution in Italian (Carminati, 2002) in a Latin American variety of Spanish. We then examined the application of this strategy by Spanish heritage speakers of the same dialect who were largely English dominant. Forty-five monolingual speakers of Mexican Spanish and 28 Spanish heritage speakers of Mexican descent read sentences in which null and overt subject pronouns were biased for and against expected antecedent biases. Our results suggest that Mexican monolinguals display distinct antecedent biases for null and overt pronouns. Furthermore, the Spanish heritage speakers, though not monolingual-like, did not violate discourse constraints on the resolution of overt pronouns, contra the findings of offline research (see Keating, VanPatten & Jegerski, 2011). We discuss the findings in terms of a processing-based account.

This article is concerned with the description and analysis of Övdalian referential null subjects. A general background to Övdalian is provided together with the syntactic restrictions on the possible null subjects (wįð ‘we’ and ið ‘you’). Interestingly, these null subjects in Övdalian do not appear in the same syntactic positions. This syntactic difference leads us to the conclusion that the distribution of the two possible null subjects must be explained individually. I argue here that the syntactic restrictions indicate that null wįð requires a link to the surrounding context in order to be identified, whereas the identification of null ið seems to be dependent on the agreement affix. I build on the proposal of Koeneman (2006), and argue that affixes may have pronominal properties, proposing that this gives an explanation as to why null ið is not restricted in the same fashion as null wįð. Finally, Övdalian is discussed in a wider Germanic context, and it is shown that Övdalian is one of a small number of non-standard Germanic languages which allow referential null subjects.