| dc.description.abstract |
Alternative semantics and the associated compositional machinery has become an important part of the formal semanticist's toolbox. Beyond its origins as a tool to model the semantics of questions (Hamblin 1973) and focus (Rooth 1985), alternative semantics is now used in a myriad of ways to model phenomena at the semantics/pragmatics interface including Negative Polarity Items (Lahiri 1998, Krifka 1992, Chierchia 2013), Free Choice (Kratzer and Shimoyama 2002, Aloni 2007, Men endez-Benito 2010), Quantifi er particles crosslinguistically (Kratzer and Shimoyama 2002, Szabolcsi 2015), Disjunction (Alonso-Ovalle 2006, M. Simons 2016) and by proponents of the 'grammatical view' of scalar implicatures in conjunction with the alternative sensitive exhausitivty operator (Chierchia, Fox, and Spector 2012, Fox 2007). Since the early days of alternative semantics, there has also been considerable discussion among formal semanticists about formal aspects of the compositional system for modelling alternative semantics for focus and just how much expressive power this system needs to adequately model association with focus and associated discourse phenomena (Rooth 1985, Rooth 1992, Rooth 1996, Kratzer 1991, Wold 1996, Krifka 2001, Krifka 2006, Beck 2006, Romero and Novel 2013). This question has not been addressed to the same extent for other phenomena where alternatives have been argued to play a role in the compositional semantics. While the tools from alternative semantics have proved extremely useful in modelling the behaviour of these other grammatical phenomena, it remains an open question to what extent they rely on the same grammatical system. This is the question at the heart of this thesis: Where do the alternatives introduced by free choice items and disjunction fi t into the compositional system of alternative introduction and manipulation underlying focus and questions? How do they interact? This thesis contributes two case studies that address this issue from a crosslinguistic perspective.
The first case study looks at alternatives introduced by disjunction in Yoruba disjunctive questions. In Yoruba, a Niger-Congo language, polar and alternative disjunctive questions are disambiguated via a syntactic and morphological focus fronting strategy that expresses exhaustive focus elsewhere in the language. I argue that the interpretation of alternative questions in Yoruba arises via the introduction of alternatives by disjunction which are operated on, first by an alternative sensitive maximality operator responsible for the exhaustivity inference observed with Yoruba ex-situ focus, and subsequently by an alternative sensitive Q-operator. The way in which these different alternative evaluating operators must both associate with a single alternative-introducing element is evidence that all of these elements (focus, questions, exhaustive inferences) employ the same kind of alternatives and, furthermore, that the grammatical system responsible for generating and manipulating alternatives must have the power to selectively evaluate alternatives.
The second case study looks at a Free Choice Item in Samoan, a Polynesian language. In Samoan, the determiner so'o se is morphologically composed of a non-specifi c determiner se, and a particle 'o, which has been argued in previous work to mark the introduction of alternatives. The use of so'o se gives rise to a universal free choice interpretation and shares a similar restricted distribution with other universal free choice items crosslinguistically. I argue that the free choice interpretation and restricted distribution of so'o se in Samoan comes about through a semantic composition employing alternatives and their interaction with two covert alternative evaluating operators: an exhaustivity-contributing operator as well as a universal quantifier over alternatives, like in the proposal by Men endez-Benito 2010 for the Spanish FCI cualquier. As with the disjunctive questions in the previous chapter, the account requires
adopting a view of alternatives under which they are able to pass on alternatives,
and able to selectively bind distinguished variables. On the other hand, puzzling data from a lack of intervention e ffects with so'o se complicates the picture, suggesting that covert movement of the alternative-introducing FCI occurs, at least in some cases.
Overall, this thesis argues for a view of alternative evaluating operators as a
single compositional mechanism available in natural language grammar that is at
work across various phenomena including focus, questions, disjunction and the
derivation of free choice interpretations. Furthermore, it highlights an interesting
crosslinguistic regularity concerning the areas in which alternatives come to be
used. |
en |
