Against severing the external argument from its verb
Hubert Haider
November 2020

The hypothesis that the subject of a (transitive) verb receives its theta-role externally rather than by the lexical verbal head merits being contested. It is wrong. The main reason of failure is the SVO bias. The hypothesis demonstrably misses the mark for languages with head-final VPs. Data from an OV-based language like German are sufficient for rebutting the representativeness of the original data basis. When the hypothesis and its predictions are systematically confronted with data from such a language, it fails. The results have immediate consequences for current claims within the Minimalist Program, with "little-v" as the theta-assigner of such (transitive) subjects. Crucially, the subject position is VP-external only in [S[VO]] languages. It is one of the type-defining properties. In SOV and also in VSO languages, a subject stays in its VP-internal base position (unless it is wh-moved). As for little v, the fact that complex head-initial VPs contain an empty V position is a derivable property. However, 'little v' as an empty verb with inherent grammatical properties is a readily dispensable concept. The obligatory, VP-external structural subject position is a predictable and derivable property of [S[VO]] languages. It is absent in SOV and VSO languages. There is neither empirical need nor theoretical necessity for an idiosyncratic "little v" or a universal VP-external functional subject position ("EPP").
Format: [ pdf ]
Reference: lingbuzz/005467
(please use that when you cite this article)
Published in:
keywords: little v, structural subject position, epp, external theta role for subjects, svo, sov, head-final, head-initial, idioms, collocations, adverb placement
previous versions: v7 [October 2020]
v6 [October 2020]
v5 [October 2020]
v4 [October 2020]
v3 [September 2020]
v2 [September 2020]
v1 [September 2020]
Downloaded:358 times


[ edit this article | back to article list ]