A major semantic distinction obtaining in phrases and clauses is that between modifiers and arguments. While arguments are inherent to the meaning of the head of the phrase or clause, modifiers merely supplement the head with additional information. Typical examples of argument-taking heads are verbs and adpositions, but also relational nouns such as kinship terms and body part terms. Typical examples of modifier-taking heads are non-relational nouns, such as those denoting concrete, inanimate objects, like ‘pot’ or ‘pen’.
This dissertation investigates to what extent the modifier/argument opposition constrains the morphosyntactic expression of possessive NPs, adpositional phrases and verbal main clauses. Using data of 64 different languages from all over the world, it is shown that the modifier/argument distinction strongly correlates with four typological parameters: locus of marking, the referentiality of person marking, the formal bondedness of person marking, and identity of marking. The first three parameters apply to possessive NPs: it is shown that possessive modifiers, as functionally optional enrichments of the head, are more likely to be the locus of morphosyntactic marking than possessive arguments. Also, possessive modifiers tend to be expressed by more referential and more formally independent person markers than possessive arguments. The final parameter applies beyond the domain of the NP: it is shown that if a language uses the same morphosyntactic form to mark modifiers in phrases and arguments in clauses, it also uses this form to mark arguments in phrases. Together, these findings demonstrate the typological relevance of the semantic opposition between modifiers and arguments, in the possessive NP and beyond.