A kind of case marking—termed variously active, active-neutral, active-inactive, active-static, stative-active, agentive, agent-patient, split S, and split intransitive—is shown to be less arbitrary than is sometimes assumed. Its semantic bases can be missed if sought only in immediate one-to-one correspondences between meaning and form. Case systems of this kind are often the products of successive diachronic developments, each individually motivated. Several factors can obscure the motivations, including not only crosslinguistic differences in detail, but also shifts of defining features over time, grammaticization, and lexicalization. To explain why these case systems have the shapes they do, we must appreciate both the diversity of features that can underlie them and the dynamic processes that mold them.