<p>Sociological theory, as represented in the two main U.S. theory journals, has been diverted from the project of generalized explanation by a propensity for over-refinement, a preference for small-scale conceptual innovations linked to specific and sometimes quite obscure cases, a tendency to focus on heroes of theory rather than their generalizable ideas, and too much attention to the flux and instability of social life. These conclusions are based on a content analysis of 445 journal articles over a recent 11-year period. As an alternative, I propose focusing on high-leverage sociological concepts, those that have explanatory power across multiple settings, subfields, and levels of analysis, as well as the variables that are most important for conditioning their explanatory force. I connect the project with a recessive theme in the work of the mid-20th century sociologist and theorist Robert K. Merton.</p>