On the role of syntactic locality in morphological processes: the case of (Greek) derived nominals 1 Artemis Alexiadou University of Stuttgart artemis@ifla.uni-stuttgart.de 1. Introduction A certain amount of consensus exists that the generalization in (1) holds across languages (though expressed from different theoretical viewpoints, see Grimshaw 1990, Bierwisch 1989, Borer 2001, to appear, Alexiadou & Grimshaw to appear among others): 2 (1) Derived nouns that have argument structure inherit this in some form from their verbal source 3 Under a specific understanding of (1), which I follow here, (1) basically says that in order for a noun to have argument structure (AS) this must have been a verb at some point in its derivational history. This suggests a very concrete relationship between morphology and the presence of AS. In particular, it suggests that in languages with verbalizing morphology, nominalizing morphology should appear at the outside of the verbalizing markers and these derived nominals should always (i) bear meanings related to their verbal source and (ii) have AS. 1 I would like to thank Hagit Borer, Heidi Harley, Hans Kamp, Florian Schäfer, two anonymous reviewers as well as the participants at the workshop on ''QP structure, Nominalizations and the role of DP'' in December 2005 in Saarbrücken for their comments. Special thanks to the editors of this volume whose comments greatly improved the readability of the paper. The idea to look at the different nominal derivational patterns of Greek grew out of a seminar on nominalizations at the graduate seminar at the University of Crete in May 2005. I would like to thank Elena Anagnostopoulou and the participants for their input and their suggestions. This work was supported by a DFG grant to the project B1: The formation and interpretation of derived nominals, as part of the Collaborative Research Center 732 Incremental Specification in Context. 2 But cf. Ehrich & Rapp (2000). 3 Note that (1) does not refer to nouns expressing kinship terms and body parts, which are taken to be inalienable possessor constructions. As I argued in Alexiadou (2003), such nouns also license arguments, the inalienable possessor being then an argument of the possessed noun.