Is US president Donald Trump a threat to democracy? Alerting against his manifold transgressions of democratic norms, many comparative political scientists have thought so. Their practical worries, however, have been inconsistent with prevalent theories of democratic stability. As careful examination shows, his main democratic norm violations have been discursive, and they have revealed him to be, not an ideological enemy of democracy, but a self-centered actor without deep democratic commitments. None of this should ring democratic alarm bells. But it does. As I suggest, Donald Trump has been conducting a kind of sociological "breaching experiment" on the political science community which has exposed a remarkable divergence between our main theories of democratic stability (which focus on structures, political behavior, and self-interest) and our tacit convictions (about the causal relevance of actors, political language, and normative commitments).