Pragmatics of the Complex Dp in Ancient Greek

In Classical Greek, complex determiner phrases may be formed in various ways. From a semantic point of view, the different formal patterns seem not to be associated with differences in meaning; in traditional grammars they are all claimed to be equally grammatical and equally definite. The distribution of the varieties of complex DP in discourse has not been studied in detail. In this paper I will discuss the combination of a noun with a modifying adjective. I will investigate the various patterns from semantic, pragmatic and, to a certain extent, syntactic points of view. I conclude that the possible configurations are semantically equivalent, but sensitive to information structure, both within the clause and in the larger context. 1 Definiteness in Classical Greek The definite determiner of Classical Greek developed from the well-known Indo-European demonstrative pronoun *so-/to-. This process of semantic and pragmatic function change has been documented for many other languages, e.g. German and the Romance languages (Lyons 1999). In Classical Greek, this development took place between the writing down of the Homeric epics, the Iliad and the Odyssey (ca. 700 BCE) and the beginning of the Classical period (500-300 BCE). There are some indications of an emerging definite determiner already in Homer, but it is only in the prose literature of the fifth and fourth centuries that we may observe the full extension of the system. The definite determiner attaches to its NP as a proclitic and agrees with it in number and case. Clitics of different types, such as sentence particles, may intervene between the determiner and its noun complement. In the case of a complex NP, there are several possible ways of attaching the determiner (1a-e):