!" In this paper, I argue that the function of the socalled 3rd person singular suffix, #, 1 is not to mark “3rd person” but to mark only possessum in Turkish. In other words, I propose that Possession, or “Possessedness”, must be an independent grammatical category in Turkish. I will frame my analysis in a processbased approach to morphology in which affixes are not lexical items but rule elements belonging to morphology, as in Zwicky 1986, Anderson 1992, Aronoff 1994, Beard 1995, Pounder 2000, and Kunduracı 2013. My analysis will also follow the assumption that morphology is an autonomous component, separate from both lexicon and syntax (e.g. Zwicky 1986, Di Sciullo & Williams 1987, Aronoff 1994, Beard 1995, Göksel 1998, 2007, Pounder 2000, 2006, Sadock 2012, Kunduracı 2013). In section 2, I will introduce Turkish possessive suffixes in possessive constructions, and will present certain challenges to the analysis of # as the 3SG suffix. In section 3, I will propose that the Turkish # marks possessedness without person information in contrast to traditional analyses. Next, in section 4, I will contrast Possessedness with Person as two separate categories. In section 5, I propose the structure for the formation of possessive constructions in Turkish, including the suffixation of #. In section 6, I will show that not all possessive constructions involve agreement. Finally, section 7 will draw conclusions regarding Possessedness, Person and agreement in Turkish possessive constructions. !"#$ % &’ ’’’’( ’’ In this section, we will see what possessive constructions (“PC”s) in Turkish look like and why the traditional assumption that # marks the 3rd person is problematic. Turkish PCs comprise a modifier NP with the genitive suffix, ()#, as the possessor, and a head NP with one of the possessive suffixes, (#)$, (#), or #, as the possessum. I illustrate these in (1): 1 # surfaces as %& %& %& %’& %& %& %& %’, and as # before case suffixes.