Free inversion in Old High German and Cimbrian. On the status of tho/da as CP-expletives and their connection with pro-drop

In this paper I compare Old High German and the Germanic dialect Cimbrian, two languages which share the availability of free inversion in co-occurrence with an expletive-like element tho and da ‘there’, and I show that they share striking similarities which follow from their pro-drop nature. Tho is typically analysed as a narrative/discourse-continuative marker (Axel 2007, Fuß & Hinterhölzl 2019) appearing in CP or in TP – an account which does not make sense though of the fact that its distribution is restricted to certain verb types (typically unaccusatives, verbs of saying) and constructions (such as passives). In this paper I reconsider the possibility that tho and free inversion are to be connected to the availability of pro-drop in Old High German (cf. Haeberli 2001 for this idea for Old English) and I show that if we apply a Topic-matching analysis for the licensing of null subjects (Frascarelli 2007, 2018) to the Old High German data we are able to solve the problems of Haeberli’s account discussed in Axel (2007). These problems all follow from the assumption that tho lexicalises a there-type expletive appearing in Spec,TP which goes against the evidence for tho. In my alternative account I show that in Old High German free inversion involves the presence of an overt or silent expletive tho, a locative argument selected by the lexical verb which can be promoted to an expletive of the TopicP position in the left periphery in sentences in which the DP subject has not moved out of the VP (cf. Tortora 2001 on free inversion in Italian) and whose function is to satisfy the EPP feature associated with the Topic-criterion needed for the licensing of pro in Spec,TP (Frascarelli 2007, 2018). The availability of free inversion was lost in the history of German due to the loss of pro-drop but it is still available in the Cimbrian dialect spoken in the village of Luserna. In Cimbrian free inversion obligatory involves overt da, a CP expletive, enclitic to the finite verb or the complementiser and whose function is purely formal, i.e. licensing pro in TP, as proposed by Bidese & Tomaselli (2018). I will suggest that the two elements are connected and that Cimbrian da is not an innovation but results from a grammaticalisation process of tho according to which the locative expletive develops from a maximal category with discourse properties (tho) into a head (da) with functional/grammatical status (van Gelderen’s 2010 Head Preference Principle).