Agreeing possessors and the theory of case

AGREEING POSSESSORS AND THE THEORY OF CASE Ion Giurgea* Abstract: I argue that the so-called “possessive adjectives” are not really adjectives, but pronouns (D-projections). Agreement features on possessors do not compete with genitive-marking, as it seems at first sight: they never attach directly to DP, but rather to KP (i.e. to a projection of Case), as shown by the obligatory presence of a possessive suffix (which I analyze as K) separating the root from the agreement morpheme. I argue that this explains why the unvalued -features of the possessor do not agree with the inherent -features (I show that some agreeing possessors do have inherent -features): concord applies inside a DP-phase, and given that DP is a phase, the features attached to K above DP belong to the phase of the possessee rather than to the phase of the possessor. Further evidence for this proposal comes from agreeing genitive markers attached at the phrasal level and from agreeing markers attached to a genitive morpheme. I then discuss the implications of this analysis for the theory of structural case: given that agreeing possessors can represent structural case, but their K must be specified as possessive from the beginning of the derivation, structural case cannot be considered to be unvalued case. Moreover, the fact that case concord is often found among agreeing possessors also shows that one and the same K head can have structural case in need of licensing and an unvalued Case feature valued by concord. Finally, I propose an account for agreeing possessors which seem to be doubly marked, both by a case ending or possessive suffix + agreement (at the word level) and by a preposed agreement marker attached at the phrasal level. Keywords: agreeing possessors, case theory, structural case, concord, genitive case 1. Introduction: The issues raised by agreeing possessors This paper† addresses the implications of the phenomenon of agreeing possessors for the theory of case in general, and for the analysis of genitive case in particular. I use the term “agreeing possessors” for those pronominal and nominal forms which are functionally equivalent to genitives, but are morphologically distinguished from genitives by being marked with agreement features copied from the head noun (the “possessee”). These forms are traditionally labelled “possessive adjectives”, or, when they occupy the D or SpecDP position of the embedding DP, “possessive determiners”; when they are based on pronominal roots, they are also known as “possessive pronouns”. The label possessive pronoun is more familiar, because many Indo-European languages only have agreeing possessors based on pronouns. But agreeing possessors can also be built on nouns, as is the case in Slavonic languages. The existence of a functional competition between agreeing possessors and genitives is well-known. The following examples show that agreeing possessors have the interpretation of genitives (complements of deverbal nouns, arguments of non-derived relational nouns, modifiers expressing possession or another contextually established * Institutul de Lingvistică “Iorgu Iordan – Al. Rosetti” and Universität Konstanz, . The research in this paper has been supported by the Sectorial Operational Program Human Resources Development (SOP HRD), financed from the European Social Fund and by the Romanian Government under the contract number SOP HRD/89/1.5/S/59758. † Io n G i ur ge a 6 relation) and sometimes block the realization of the genitive of the corresponding pronoun1: (1) (2) (3) Complement of deverbal noun: a. arestarea {Mariei / noastră/ *nouă} de către jandarmi (Rom.) arresting-the. F SG Maria-GEN/DAT / our.F SG / us.DAT by gendarmes b. {l’arrestation {de Marie / *de nous} / par les gendarmes (Fr.) the arresting of Marie of us by the gendarmes b′. notre arrestation par les gendarmes our.SG arresting by the gendarmes Argument of relational noun: a. fratele {Mariei / meu / *mie (Rom.) brother-the.M SG Maria.GEN/DAT / my.M SG / me.DAT b. le frère {de Marie / * de moi} (Fr.) the brother of Marie / of me b′. mon frère my.M SG brother Modifier (possession or other contextual relation)2 : a. casa {Mariei / mea / *mie} (Rom.) house-the.F SG Maria.GEN/DAT / my.F SG / me.SG DAT b. la maison {de Pierre / *de lui} / (Fr.) the house of Pierre / of him / b′. sa maison his/her.F SG house Regarding positions, there is an array of possibilities across languages, going from total disjointness between agreeing possessors and (inflectional or prepositional) genitives to total overlapping; where the positions are different, it is always agreeing possessors that occupy positions more to the left, presumably in the functional domain of the noun phrase: (i) agreeing possessors always occupy positions where genitives are excluded; (4) a. b. 1 {mon / *de Marie} livre my.M SG / of Marie book le {mien /*de Marie} the my.MSG / of Marie (Fr.) This only holds for pronominal possessors; for nominal agreeing possessors, see Partee and Borschev (2003) on the interpretations available for the agreeing possessor and the corresponding genitive. 2 The availability of various contextual relations for the genitive modifier was pointed out by Williams (1981), who enumerates the following possible readings for the phrase John’s car (of course, the list remains open): (i) a. the car that John owns b. the car that John is renting c. the car that John has to paint d. the car that John has to find e. the car that John prefers Agreeing possessors and the theory of case c. (5) a. b. le livre {de Marie /*mon} the book of Marie my Petin stul (Partee and Borschev 2003: 23-24) Petja-POSS M SG chair-M SG stul Peti chair-M SG Petja-GEN SG ‘Petja’s chair / a chair of Petja’s’ 7 (Rus.) (ii) some agreeing possessors (namely weak forms) occupy positions where genitives are excluded, others occupy the same positions as genitives; (6) a. b. {mi / *de Luís} libro my.SG / of Luis book ‘my book’ el libro este {mío / de Luís} the book this my.M SG / of Luis ‘this book of mine / of Luis(’s)’ (Sp.) (iii) agreeing possessors are allowed in a sub-set of the environments in which (some type of) genitives are allowed3. (7) a. b. {meine / Johannes} Schwester my-F SG / Johann’s sister eine Schwester {*mein(e)} / {Johannes} a sister my-F SG) / Johann’s (Germ.) The fact that the agreeing possessor occupies in (7)a) the same Spec position as the genitive, rather than a head position, is shown by the possibility of coordination with a genitive: (8) Sie treffen sich in [deiner und Karls] Lieblingskneipe (Olsen 1989) they meet REFL in your.F SG DAT and Karl.GEN favourite-bar.F SG (iv) agreeing possessors are allowed in all the environments in which some type of genitives are allowed; this is the case of Romanian, where both inflectional genitives and agreeing possessors immediately follow either the suffixal definite article, or the so-called “genitival article” al, which is a genitival marker (...truncated)