Can current methods of pathonormal inference tell us anything about modularity?

This commentary on Muller's target article is designed to help bridge the gap between method and theory in neurobiological approaches to language. We focus on pathonormal inference as reflective of broader conceptual issues concerning universality and modularity. Our commentary directs attention to how current methodological and analytic procedures are less than optimal and minimize possibilities for reaching theoretical consensus. We make suggestions about alternative analytic strategies that provide a direction for future scientific progress.