Focusing on Darwinism, David Hull argues that the protean character of conceptual systems is explained by their nature as historical entities which evolve. If they evolve as biological species do, Hull argues, then they cannot have an "essence" — a set of tenets that all and only instances of the conceptual system has throughout all time. There are no tenets a scientific research program must retain to count as an instance of a particular program. I advance two considerations against this view. First, research programs require a critical cohesiveness among their tenets to inspire and guide research. Second, it is the function of such programs to guide the search for answers to families of questions in a particular domain in a particular spirit. These factors dictate that conceptual systems must retain certain key tenets. This re-emergence of a sort of essentialism does not bar the evolution of conceptual systems, provided we recognize that there are patterns of evolution other than the one Hull considers (anagenesis). It also implies that conceptual systems simply evolve differently than species do. I defend this position by illustrating two episodes of conceptual evolution: the dispute between William Bateson and the British biometricians over discontinuous evolution, and the formation of Neo-Lamarckism in 19th century America.