Workshop on Inversion and Verb Movement

The V2 conspiracy revisited

Thórhallur Eythórsson
University of Iceland, NORMS

In this paper I address the following questions concerning V2 in Germanic: 

(a)       How old is the V2 phenomenon?  

(b)       What is the earliest distribution of V2?  

(c)       How did V2 develop? 

(d)       To what extent has V2 been lost?   

The answers to these questions are based on an extensive examination of the relevant empirical evidence in Gothic, Old English, Old Norse and other old (and modern) Germanic languages. Following up on previous work (e.g. Eythórsson 1995, 1996, 2001), I argue that already at the earliest stage of Germanic (Gothic) there is evidence supporting the assumption of V2 in well-defined types of main clauses: wh-questions and with certain fronted phrases in clause-initial position. This type of V2 arguably involves the same syntactic operation ("V-to-C”) as some subtypes of V1 (itself not a homogenous phenomenon), e.g. intitial verbs in questions,  commands and negated contexts, as well as the so-called narrative inversion. 

Moreover, I propose that V2 was extended from this limited context to other types of main clauses, i.e. those beginning in a subject and a fronted complement. This happened in North Germanic and continental West Germanic (whereas the status of Old English as a full V2 language is less clear). In this connection, I focus on both syntactic and phonological evidence which did not receive full attention in previous work. Finally, I claim that the generalization of V2 in embedded clauses in North Germanic involved the extension of the pattern found in main clauses. In addition, I investigate cases of loss of V2, for example in the history of Mainland Scandinavian, in order to establish what such cases can reveal about the nature of this phenomenon. 

In the light of the above considerations, I propose that V2 in Germanic developed in three stages: 

(i) wh-questions and focussed phrases (common Germanic) 

(ii) other types of main clauses (North Germanic, continental West Germanic) 

(iii) embedded clauses (symmetric V2) (North Germanic)   

According to the historical scenario defended here, the distribution of V2 in Old Germanic is not the outcome of a single rule, but rather of various independent rules, both syntactic and phonological, conspiring to target particular structural positions. In this sense it is legitimate to speak of a "V2 conspiracy”, as in Weerman (1989), although the specifics of that account are not adopted here.