Accounts of who leads language change are abundant, and often play an explanatory role in models of how language changes begin and spread. In this talk, I argue that the notion of a “leader of language change” is not on a strong footing empirically. Using data from six ongoing sound changes in white Philadelphian English, I show that although young women generally are most advanced in all of the changes, there is no consistency in terms of which particular young women lead each change. In other words, we cannot predict from a speaker’s relative position in one change whether they are innovative or conservative in other changes. I contend that this type of covariation failure is unexpected and may be difficult to reconcile with many models of language change.