Huang, Y.
Harvard University
It has been widely acknowledged that there-insertion is incompatible with unergative verbs. Therefore, it has been established as a syntactic test for unaccusativity. This paper presents the results from two judgment experiments to argue that using there-insertion as an unaccusativity test is empirically ungrounded. The first study shows that although there is a statistical difference between unaccusative and unergative verbs w.r.t. their acceptability in there-insertion construction (p-value < 0.05), the significance is actually caused by one subgroup - Verbs of Existence. In the second study, participants were forced to choose a better sentence from an unaccusative-unergative sentence pair. The results show that there is a preference for unaccusative verbs in there-insertion condition (65% vs. 35%). However, the choice is not categorical. There are much variation among sentences and speakers. No consistent preference is found for any subcategories of unaccusative verbs. These results are problematic for using there-insertion as a test for unaccusativity, which assumes a clear cut syntactic distinction between two classes of intransitive verbs and a consistent choice for unaccuative verbs over unergative verbs. Rather, the results are gradient and heavily influenced by individual verb meaning.