He then said..: Understudied deviations from V2 in Early Germanic

ABSTRACT This paper discusses a V3-pattern in Early Germanic that has so far not been considered independently. In this construction, a clause-initial XP is followed by the adverbial element OHG do/OE þa/OS tho (lit. ‘then’), which is directly followed by the finite verb. Based on a pilot study of the OHG translation of Tatian’s gospel harmony and the OE Blickling Homilies, it is shown that the pattern exhibits slightly different properties in OE and OHG. In OHG, the element preceding do is usually a pronominal shifting topic, while in OE, the clause-initial XP may also be a full DP that is either a shifting topic or a continuing topic. To account for these differences between OE and OHG, we argue that OE þa is first-merged as the head of a clause-medial projection that serves to mark the boundary between the topic and the focus domain. In contrast, OHG do (and OS tho) is a topic marker that is either part of the fronted shifting topic, or base-generated as a head in the left clausal periphery. As to its internal syntax, we propose a grammaticalization path for do/þa/tho in which a demonstrative adverb first turns into an adverbial discourse marker that may also serve expletive functions before it eventually grammaticalizes into the topic particle addressed in this paper.