See Section 2.1. 2 / Possessive Descriptions the Perspective Paradox Syntactic Structure Spec-of-dp Possessive Compound Possessive Descriptions: Lexical and Extrinsic Possession

Preface I completed my dissertation in 1991, and this CSLI edition presents that work essentially unchanged. However, an anonymous CSLI reviewer raised a number of excellent points that led me to add some explanations and some further discussion, and I have made a small number of minor changes here and there as well. Thanks again to my dissertation committee: William A. Ladusaw (the chair), Peter Lasersohn, and James McCloskey; to Robert C. Moore and my former colleagues at SRI, where I wrote most of the dissertation; and to all of those mentioned in the acknowledgements section of the dissertation. Thanks also to the Linguistics Board at the for providing funds for distribution of the dissertation manuscript. I am grateful to an anonymous reviewer and to Tony Gee for help preparing the CSLI version. I would like to mention three of the papers that grew out of my dis-sertation research. The investigation into the semantics of relational nouns in chapter 2 lead to collaborative work with David Dowty, a part of which has been reported in our 1993 NELS paper 'Non-verbal thematic proto-roles' (A. Schafer, ed., proceedings of NELS 23, GSLA, Amherst); the treatment of donkey anaphora and the proportion problem in chapter 4 gave rise to 'A pre-suppositional account of proportional ambiguity', to appear in Natural Language Semantics; and a recent manuscript proposes an explanation for so-called double genitives (e.g., a friend of John's), a topic which completely baffled me at the time I wrote the dissertation. These papers, as well as a pro-log implementation of the fragments of the first three chapters of the disser-tation, are available by contacting me at [email protected]. Finally, as before, this dissertation is dedicated to Geoffrey K. Pullum, who was so important to my early professional life. 0 Introduction Possessive descriptions are extremely common. On the front page of today's New York Times, ¤ for instance, I counted 23 instances of the possessive construction , distributed in such a way that one out of five sentences contained at least one possessive. As a second kind of example, children acquire the possessive early in the two-word stage, so that by the age of 2 possessives can account for up to twenty percent of a child's productions. ¥ In other words, possessives are a basic and important part of the language. Nevertheless, possessives have not been extensively studied by genera-tive linguists in general and semanticists in …