This paper defends what we call the Straightforward View of conceptual engineering. Canonical examples of conceptual engineering are simply cases in which a proposition is endorsed or denied at one time, and this endorsement or denial is subsequently retracted; and this phenomenon is continuous with belief- and theory-change in science, history, and other areas of investigation. Put simply, conceptual engineering is inquiry. This view is consistent with standard characterisations of conceptual engineering as potentially revisionary, as ameliorative, and as a matter of assessing and improving representational devices. We reject the alternative Meaning Change View of the canonical examples, according to which they involve changing the meaning of some term, on the grounds that it is unmotivated. We also defend a novel version of the implementation objection to the Meaning Change View: the proponent of Meaning Change’s best strategy for defending the idea that the conceptual engineering is ameliorative makes it impossible that meaning be changed.