Kazakh esh-words and negative concord

Abstract

The goal of this thesis is twofold. First, I present compelling evidence that Esh-words in Kazakh pattern with Strict Negative Concord Items (SNCIs) of the variety found in Slavic languages (e.g. Russian) rather than alleged Negative Polarity Items in Altaic languages like Turkish. Secondly, I propose a novel perspective on SNC which presents significant advantages over existing ones (c.f. Zeijlstra 2004). The most explicit account of SNC languages (Zeijlstra 2004), makes the undesirable assumption that the sentential negative markers in them are just another SNCI licensed by a silent negative operator, thus failing to explain why its presence is obligatory to license all other SNCI. This proposal, instead, focusing on the absence of negative quantifiers (like no NP) in these languages, argues that syntactically local combinations of SNCI phrases with sentential negation is a device to express no NP. The proposal explains why Esh-words must occur with negation: Combining an existential determiner with a non-antimorphic function does not result in a negative determiner. Similarly, intervening quantifiers of any sort disrupt the equivalence with no, therefore the proposal makes two additional desirable predictions: i) Strict NC items are sensitive to intervention effects. ii) Strict NC items must occur in the same clause as the negation licensing them.