The requirement that candidates in studio-based or practice-led higher degrees by research should formulate a research question has been found to be problematic by some writers. The present article argues that this stance, particularly as it is articulated by proponents of the influential category of "performative research" (Haseman, 2006, 2007), is ill-founded and that formulating a research question correctly can instead assist the creative project and prevent it separating into two separate enterprises: one practical and the other theoretical. Drawing on the principles of General Semantics as formulated by Alfred Korzybski, it suggests ways in which the relation of the question to ontological considerations can affect its usefulness for studio-based research.