| This thesis aims at obtaining a unified account of Negative Polarities Items (NPIs) and Negative Concord Items (NCIs), which further divide into two types, strong and weak. Strong NCIs refer to the type of NCIs whose distribution is limited to the environment where they have a clausemate sentential negation marker. Weak NCIs differ from strong kinds in that they exhibit a preverbal-postverbal asymmetry with respect to negativity: preverbal instances are able negate sentences by themselves while postverbal ones require the presence of a sentential negation marker. In addition, weak NCIs bear another distinctive feature: they can produce an existential reading in a subset of the contexts where NPIs would be used in languages like English. The main questions addressed in this thesis are (i) why strong NCIs require a clausemate sentential negation marker, a non-trivial question that arises from the fact that strong NCIs can appear alone in elliptical answers, (ii) how the structure that contains a strong NCI and a clausemate sentential negation marker yields a single negation reading, (iii) why weak NCIs exhibit the above-mentioned preverbal-postverbal asymmetry and apparent semantic ambiguity (between negative and existential readings), and (iv) what relation holds, if any, between strong NCIs, weak NCIs and NPIs. To answer these questions, following Watanabe (2004), I will first conduct a cross-linguistic examination of the morphosyntactic ingredients of NPIs and the two kinds of NCIs. The result shows that NPIs and strong NCIs share a focus element and an indefinite, the difference being that the latter contain a morpheme related to a sentential negation marker, and that weak NCIs are built upon a strong NCI with the addition of a focus- and negation-related morphemes. With this morphosyntactic analysis, the above four questions will be answered in tandem with the proposal that Agree is a feature-sharing operation a la Pesetsky and Torrego (2004). |
PDF Full Text Request