This paper describes a novel *local > local ditransitive person restriction in Kipsigis (Kalenjin; Nilo-Saharan) and argues that this restriction must be morphological. Ditransitive person restrictions are often accounted for syntactically, as done for various forms of the Person-Case Constraint (PCC). However, the Kipsigis pattern differs from traditional PCC effects in two ways. First, only local > local object combinations are ungrammatical -- an unattested PCC pattern. Second, *local > local is tied to morphological competition; when competition is removed, the restriction is obviated. Given this, the Kipsigis facts must be accounted for morphologically, which suggests that the underlying cause of ditransitive person restrictions can vary cross-linguistically: stemming in some cases from the syntax and in others from the morphology.