Abstract. — Nearly every fan of conceptual engineering is familiar with ‘Strawson’s problem’. This worry, first suggested by Strawson’s critique of Carnapian explication, concerns the possibility that our conceptual interventions might problematically ‘change the subject’, thereby disrupting inquiry and potentially allowing important philosophical problems to be ‘solved’ via stipulation. A complete theory of conceptual engineering needs to address this worry in some manner or other, if only to dismiss it — in other words, a complete theory of conceptual engineering needs to specify the limits (if any) on permissible conceptual revision. My talk focuses on a variety of cases that I hold to pose a challenge to developing a successful account of the limits of revision: cases involving multiple successors to an original pre-engineering concept. I’ll argue that many prima facie plausible approaches to specifying these limits have the undesirable implication that a very large proportion of multiple-successor engineering proposals — including numerous examples from 20th-century philosophy — are impermissible.