GENERIC AND INDEFINITE NULL OBJECTS Generic and indefinite null objects

Věra Dvořáková, Sam Al Khatib, Pavel Caha, Chris Collins, František Daneš, Manfred Krifka, Petr Karlík, Ivona Kučerová, Idan Landau, Alec Marantz, Andrew Nevins, Marta Ruda (+10 others)
2017 unpublished
This thesis is concerned with the syntactic and syntactico-semantic properties of two types of non-overt internal arguments: the so-called generic null objects (GNO), as in Lars von Trier's movies always shock , and indefinite null objects (INO) as in John reads / is reading . In addition to the known data on GNO and INO, coming mainly from English, Italian, and French, it utilizes data from Czech, a Slavic language with rich inflectional morphology, which enables a novel perspective on how
more » ... spective on how these invisible objects are derived in language. I argue against the predominant view that GNO are syntactically pronouns (Rizzi 1986; Authier 1992a,b), consisting of a D-feature and/or a set of ϕ-features (Landau 2010), and possibly receiving case. Evidence is provided that albeit syntactically represented, GNO consist of a single syntactic node, little n, bearing just the interpretable gender feature, but no number or person features. Rather than pronouns with a fully developed nominal functional projection, GNO should be conceived as conceptually impoverished nouns (i.e. not containing any root), whose only semantic contribution is the one associated with an interpretable gender feature on n, namely the property of being Persona or Female Persona. Such nominal heads introduce a variable that gets bound by a generic operator (GEN), along the lines of Heim 1982 and Krifka et al. 1995. The advantage of the proposed analysis over the existing ones is that it can systematically account for both genericity and humanness of GNO without having to stipulate them as separate semantic features associated with ii GNO. Moreover, it is supported by the existence of the same gender-marked silent n in persona-denoting nominalized adjectives, outside of the context of generic statements. For INO, I adopt the general view that they are derived by a locally applying existential quantifier but I refute the theories that locate this operation in the lexicon, either as a rule operating on a given transitive predicate (Bresnan 1978, Dowty 1978, a.o.), or by positing two different predicates, a transitive and an intransitive one (Fodor and Fodor 1980). Instead, I analyze intransitivization as a generalized type-shifting operation on a verbalized root that is available if the merger of the little v and a root denotes a binary relation of individuals and events, i.e. if it is of type e, vt (whereby v-node is understood in the sense of Marantz 2007, 2013, as a verb-building, event-semantics introducing head, separate from Voice). I support this approach by the high productivity of constructions with INO, not limiting them to any particular lexical semantic class; by the observation that a lexicon-based approach loses the generalization about the restrictedness of intransitivization to imperfective verbs, including the syntactically derived secondary imperfectives; and by the role that context plays in licensing INO, in providing the property/kind of the entity that is being existentially quantified (which I formalize as a presuppositional condition for the intransitivization). Focusing on the second, most complex argument, I argue that the incompatibility of INO with perfectives follows from an unvalued EPP-like feature on perfective aspectual heads. Its existence is independently motivated by the quantificational requirements of perfective verbs in Czech, expressed in terms of a syntactic argument type or a quantificational prefix that they have to merge with. I argue that the perfectivity feature (Q Pf ) requires the movement of the direct object of monotransitive verbs out of Spec,vP to Spec,Asp Q , which is something INO cannot perform due to their non-presence in syntax. I further demonstrate that INO are not isolated in their inability to satisfy the perfectivity feature, being accompanied by existentially quantified bare plural and mass nouns, which, although syntactically represented, cannot be interpreted outside of a vP, in higher layers of a verbal functional projection. To support that INO's inaptness for perfective constructions has nothing to do with their phonological nullity, I compare them to generic null objects, analyzed in the first iii part of the dissertation. The proposed analysis of GNO as syntactically represented variables that move out of vP to the restrictor of a generic quantifier, presumably via Spec,Asp Q , predicts their compatibility with perfective verbs, as confirmed by the data. iv Acknowledgments I wouldn't be where I am today if it weren't for the many people that I met along the way. One person stands out in particular -without his continuous support and unfailing patience, this dissertation would never see the light of day. And not just that. Mark Baker was present at four crucial moments in my linguistic career. He was the first person from Rutgers who I interacted with when submitting my linguistic work, and he then encouraged me during the application process. He chaired my first qualifying paper on ditransitives, which was well received and fueled my passion for generativist analysis of some of the basic syntactic constructions in my native Czech. Later on, Mark struck the initial spark for this dissertation topic. When I confidently presented him with the aspect-conditioned contrast between object-drop allowing versus object-drop disallowing nominalizations as an argument for their syntactic derivation, he simply asked what the presented contrast shows about syntax. This dissertation is a (somewhat long, I admit) answer to this innocent question. And finally, Mark agreed to be the chair of my dissertation committee, persistently following all of my steps, some of them backwards, some of them in circles, always offering a helping hand. My immense thanks go also to my two committee members, Roger Schwarzschild and Ken Safir. If it wasn't for Roger's inspirational semantic seminar on Tense and Aspect that I took part in, I don't think I would ever have dared to delve into the intricacies of Slavic aspect, the topic I had sworn during my master's studies to never work on. Roger was also the only faculty member sitting in all three committees that guided and judged my written work during my PhD studies, twice as a committee member and once as a chair, when helping to shape my second qualifying paper on secondary imperfectivization. I don't think I ever met a person with a brighter mind or a more profound ability to question anything worth questioning. v Ken Safir was my American version of Knut Tarald Taraldsen (my Norwegian teacher during my master's studies and Ken's good friend). Without Ken's insight into the framework, deep knowledge of numerous languages, and kind-hearted spirit, our department wouldn't be what it is. His ability to test predictions made by our syntactic proposals and come up with the most unthinkable of testing constructions has always amazed me -and the discussions with him have always been a great source of inspiration. My special thanks belong to my external committee member, RadekŠimík. The decision to add him to the loop in the last year of working on the dissertation was the best decision I could make. His careful, to-the-point commenting style, super-responsiveness, and his orientation in the literature make him the ideal opponent, not to mention the valuable feedback on Czech data that he provided for me. I am sure he is going to become a highly sought-after dissertation supervisor in the course of his linguistic career, and I wish him many great students, especially from among the Czech and Moravian speakers. The list of linguists that were instrumental in the origination of this thesis would be incomplete if it did not include Veneeta Dayal. Her seminar on (In)Definiteness and Genericity was one enriching experience. Veneeta's handouts, overflowing with my own notes, inspired a good deal of my thinking about the Czech data, as evident from the content of this thesis. I never thought of myself as a semanticist, but Veneeta's presentations and papers gave me a much needed insight into the utility of compositional semantics in the analysis of seemingly syntactic phenomena -and its necessity in the analysis of borderline phenomena that neither syntacticians nor semanticists want to take full responsibility for. I am grateful to numerous people that discussed with me bits and pieces of my dissertation work, and who, sometimes by pointing out a single reference or a piece of data, left their imprint in the final outcome; namely . Thanks are due to the reviewers and audiences at OLinCo 2 in Olomouc, BLS 41 at Berkeley, FASL 24 at New York University, CUNY Syntax Supper in the Fall of 2015, and NELS 48 at the University of Iceland, where I presented parts of my dissertation research and received a lot of valuable feedback. vi My stay at Rutgers University would not be what it was without my dear classmates, Teresa Torres Bustamante, Aaron Braver, Jeremy Perkins, and Todor Koev. We spent a great many moments together, whether in the classroom or on regular hikes organized by Jeremy. I am glad I can hereby proclaim that ours is one of the most successful Rutgers linguistic classes in terms of the number of PhD completing students. Speaking of Rutgers, I cannot omit thanks to my generous sleep-over providers -