Abstract Roderick Chisholm argues that in giving an account of knowledge, we must either begin with an account of what knowledge is , and proceed on that basis to identify the particular things that we know, or else start with instances of knowledge, and proceed on that basis to formulate a definition of knowledge. Either approach begs the question against the other. This is the epistemic wheel. This article responds to C hisholm's challenge. It begins with cases of knowledge attribution and builds its account from there, identifying those features that we take to be present in the cases where we have attributed knowledge and those features that seem important when we want to withhold an attribution of knowledge. The proposal does not beg the question against either particularists or methodists; it takes the best features of each view, without beginning with either, and thereby removes us from the wheel.