No Verb First
Since the pioneering work of den Besten (1977) and Anderson and Chung (1977), a vast body of literature associates specific syntactic properties with verb-first (V1) and verb-second (V2) word orders in natural languages.
I re-examine the V1-V2 divide and I propose to demonstrate that there are no V1 languages, that is I propose to reanalyse the 9.6% of human languages that are considered V1 languages as fundamentally verb-second (V2). The Central Hypothesis is that so-called V1 languages in fact involve a class of inconspicuous preverbal elements. An inconspicuous element is phonologically or morphologically weak, but behaves just like any full constituent, head or XP, at the level of syntax. At the syntactic level, therefore, V1 languages are actually verb-second. From this perspective, both V2 and V1 languages instantiate two types of word orders:
a) *Conspicuous V2 orders*: A conspicuous phrasal constituent fills the preverbal position (prototypically German or Dutch)
b) *Inconspicuous V2 orders*: an inconspicuous element fills the preverbal position, leading to the appearance of V1 (prototypically VSO, null subject languages)
Languages that seem profoundly different (V1 vs. V2) are in fact structurally similar (V2). The key variation between (a) and (b) derives from the lexical inventory of a given language. In the prototypical V1 domain, I show that languages provide a large inventory of 'inconspicuous' preverbal elements. In the prototypical V2 domain, I present a set of syntactic effects that are triggered by the sole avoidance of verb-first.