In an effort to carve a distinct place for social facts without lapsing into a holistic ontology, John Greenwood has sought to define social phenomena solely in terms of the attitudes held by the actor(s) in question. I argue that his proposal allows for the possibility of a "lone collectivity" that is (1) unpalatable in its own right and (2) incompatible with the claim that sociology is autonomous from psychology. As such, I conclude that the relevant beliefs need to be held by more than one person.