Practicing scientists' views of science recently have become a topic of interest to nature of science researchers. Using an interview protocol developed by Carey and Smith that assumes respondents' views cohere into a single belief system, we asked 15 research chemists to discuss their views of theories and experimentation. Respondents expressed a range of ideas about science during interviews, but in ways that defied assignment to a unitary, coherent belief system. Instead, scientists expressed more or less constructivist ideas depending upon the questions asked. We interpret our findings to suggest that efforts to categorize such conceptions at best oversimplify, and at worst misconstrue, people's views about science and how such ideas might develop.