Miriam Butt: the Structure of Complex Predicates in Urdu
Total Page:16
File Type:pdf, Size:1020Kb
THE STRUCTURE OF COMPLEX PREDICATES IN URDU a dissertation submitted to the department of linguistics and the committee on graduate studies of stanford university in partial fulfillment of the requirements for the degree of doctor of philosophy By Miriam Jessica Butt December, 1993 c Copyright 1993 by Miriam Jessica Butt All Rights Reserved ii I certify that I have read this dissertation and that in my opinion it is fully adequate, in scope and in quality, as a dissertation for the degree of Doctor of Philosophy. Peter Sells (Principal Advisor) I certify that I have read this dissertation and that in my opinion it is fully adequate, in scope and in quality, as a dissertation for the degree of Doctor of Philosophy. Charles A. Ferguson I certify that I have read this dissertation and that in my opinion it is fully adequate, in scope and in quality, as a dissertation for the degree of Doctor of Philosophy. Paul Kiparsky I certify that I have read this dissertation and that in my opinion it is fully adequate, in scope and in quality, as a dissertation for the degree of Doctor of Philosophy. Veneeta S. Dayal (Rutgers University) Approved for the University Committee on Graduate Studies: iii I dedicate this dissertation to my brother, Yassin Mischal Butt. iv Abstract Constructions variously described as complex predicates, compound verbs, composite pred- icates, or serial verbs range across quite an impressive number of differing expressions in differing languages. As such, it is not immediately obvious how to arrive at a consistent formulation of the differences, if there are any, among them, and how to provide a unifying analysis of complex predicates crosslinguistically. The South Asian language Urdu employs a wide variety of complex predicates. I con- duct a detailed examination of two differing Urdu complex predicates: the permissive and the Aspectual complex predicates. The Urdu permissive brings into focus the essential prob- lem complex predicates pose for theories of syntax. The permissive is a complex predicate formed by the combination of two semantic heads in the syntax (not the lexicon). These semantically distinct heads correspond to a single syntactic predicate, which may be dis- continuous. I show that this discontinuity at phrase structure does not affect the status of the permissive as a complex predicate. The problematic aspect for theories of syntax is thus the question of how to represent the fact that a complex predicate may behave both like a syntactically complex structure with respect to certain phenomena, and like a syntactically simple structure with respect to other phenomena. Within Lexical-Functional Grammar (LFG), this mismatch in semantic and syntactic information is easily represented in terms of independent levels of representation which are related to one another by a the- ory of linking. However, LFG as originally formulated does not allow for the existence of a semantically complex, but syntactically discontinous single head. Recent work intent on solving this problem in a variety of theoretical frameworks has tended towards an argument structure approach. In particular, Alsina (1993) formulates an argument structure account of Romance causatives within LFG which allows for the argument structure composition of two discontinuous heads in the syntax. I follow Alsina in proposing an analysis of complex predicate formation at argument structure. However, v rather than moving towards a progressively minimalistic and abstract argument structure, which does not explicitly contain thematic role or other semantic information, as proposed in Grimshaw (1990), S. Rosen (1989), Ritter and S. Rosen (1993), and Alsina (1993), I take up the kind of argumentation found in Van Valin (1990), for Role and Reference Grammar (RRG), and propose an elaborated argument structure based on Jackendoff’s (1990) theory of Conceptual Semantics. Urdu Aspectual complex predicates provide evidence for an elab- orated level of argument structure. An Aspectual complex predicate is well-formed only if constraints on semantic properties such as volitionality and inception/completion are met: a main verb negatively specified for one of these domains cannot combine with a light verb positively specified for the same domain. In conclusion, this dissertation presents an in-depth examination of the structure and properties of two differing Urdu complex predicates, the permissive and the Aspectual complex predicates. I formulate a unifying theory of complex predicate formation and in the process address issues concerned with argument structure, linking, and case marking. Finally, I show that the theory of complex predicates I present not only allows a successful account of Romance restructuring verbs and Japanese suru ‘do’, but can also be used as a firm base of comparison for an analysis of serial verb constructions. vi Acknowledgements Having put to paper everything I could about complex predicates within the constraints of time, space, knowledge, and wisdom inherent in a dissertation, I find that I do not want to turn my back on the subject in weary relief. Instead, I still seem to harbor a deep seated belief that a complete understanding of complex predicates is bound to yield the question to the answer of Life, the Universe, and Everything. K.P. Mohanan was my first mentor in linguistics and the first person to introduce me to complex predicates. His delight in exploring alternative, inside-out ways of looking at things left me confused at first, but ultimately taught me very much about linguistic and scientific methodology. My linguistic debt to Mohanan is vast, and I would like to thank him for his time and patience with me. When Mohanan left Stanford for warmer climes, Peter Sells inherited several Mohanan waifs and proceeded to firmly guide us through the process of writing a dissertation. The particular genius of Peter Sells for me consisted of the fact that, no matter how garbled my thoughts were, he could recognize the essence of what I was trying to get at in no time, and sort out my jumbled thoughts into an organized system. In addition to this, Peter Sells was unique in the amount of effort and care that went into the process of advising me: he was always there to respond to desperate cries for help, and always had a good joke to crack as well. Working with Peter Sells was an invaluable part of this dissertation. The dissertation could also not have been written without the input from Joan Bresnan, Charles Ferguson, Paul Kiparsky, Veneeta Srivastav, and Tom Wasow. Joan Bresnan was a member of my committee until my final year, her sabbatical year. A meeting with Joan Bresnan is representative of perhaps the most challenging aspect of my graduate school career. I would leave her office either in complete devastation, resolved not to become a linguist at all, or beaming with exultation because the intense questioning, examination, and discussion of ideas and linguistic issues at large had been so deeply satisfying and engaging. vii Charles Ferguson gave generously of his time, taught me very much about South Asian literature I was not aware of, and deceptively mildly insisted that I do things properly. This dissertation would not be as thorough without his input. Paul Kiparsky and Veneeta Dayal joined the committee as the dissertation was already well under way, but had an immediate and pronounced effect. Paul Kiparsky attacked issues of morphology and lexical semantics I had not considered before. Veneeta Dayal took a close look at my data and claims, pushed me on them steadily, and streamlined my argumentation in the process. Tom Wasow, who served on the Oral Exam committee, generously commented on my work in precise detail. His input allowed a more intelligent presentation of my material. Finally, I would like to thank Yoshiko Matsumoto for serving as the chair of my Oral Exam committee. Alex Alsina, Mary Dalrymple, Henri¨ette de Swart, and Tara Mohanan all read parts of the dissertation and provided me with detailed and insightful comments time and time again. In particular, I could not have done without the many discussions with Mary Dalrymple about the architecture of LFG, and the many exchanges of ideas with Tara Mohanan about the nature of complex predicates and the structure of Hindi or Urdu. In this respect, Alice Davison at the University of Iowa had a most stimulating effect as well. Many sections in this dissertation arose from a need to address the questions and issues she raised. I also owe many thanks to the more general linguistic community at Stanford. Many dis- cussions with Lynn Cherny, Michael Inman, Michio Isoda, Chris Manning, John McWhorter, Yo Matsumoto, and Chris Pi˜n´on kept me on my toes and provided me with interesting material to think about. Veena Dwivedi, although at the University of Massachusetts at Amherst, must also be included among this list, as she was the one person I could always most immediately engage in discussions on the structure of Hindi/Urdu. Jennifer Cole, Tracy King, and Gillian Ramchand are the fellow graduate students without whose friendship and support I could not have done. From the time Tracy King and I organized our first conference together in our second year, it was clear that we made a great team. As a team, we went on to organize another conference, and to co-author a paper. In this spirit, we might have co-authored a dissertation, except for the fact that she benightedly insisted on working on Russian. Many aspects of this dissertation have benefitted from Tracy’s input, and some of the discussion on case marking is based on joint work. Gillian Ramchand was the most delightful peer I could have had in graduate school. Whether it was over breakfast, over an A’s game, or during a session of late night TV viii watching, she was always ready to engage in a discussion of linguistic issues, or explain a particular problem I had not been quite able to come to grips with as yet.