A lot of Greek papyri, i.e., a plant-based common writing surface in ancient times, have been discovered in the sands of Egypt. These papyri come in various types and belong to different eras. Among them, documentary papyri, i.e. non-literary papyri, include contracts, petitions and letters and were written by people of various social, educational and cultural background. In my project, drawing on data from a corpus of non-literary Greek papyri from Egypt dating between Ist and -VIIIth cent. AD, I propose to investigate the behaviour of headed relative clauses (i.e. syntactic configurations corresponding to, for instance, 'the cat that the mouse chased') in relation to the sociolinguistic level of the text they occur in. In particular, I intend to test the hypothesis, which has never been applied to Greek, that there is a direct correlation between the complexity of a headed relative clause and the level of the register which the text hosting the relative clause belongs to. Complexity, however, is itself the result of different linguistic factors that make a specific configuration informatively less necessary or harder to process. Systematically examining six parameters of variation, this research will broaden our understanding of complexity in headed relative clauses and shed light on their distribution in a rather understudied period of Greek language.