Complex tool use is one of the defining characteristics of our species, and, because of the good preservation of stone tools (lithics), one of the few which can be studied on the evolutionary time scale. But a quick look at the lithics literature reveals that, although our natural human tendency is to talk about tools in general primarily in terms of their functions (e.g., hammer, knife, etc.), our stone tool typologies contain a mixture of terms relating to guessed function, manufacturing method, and shape. As such, they are entirely unsuitable as units in larger evolutionary analyses. In this paper, I argue based on the psychological literature that function is the primary concept underlying tool concepts in humans and non-human primates. Furthermore, knowing how function is represented and learned in both human and non-human primates offers insights into the reasons why current typologies are insufficient and how they could be constructed to lead to more meaningful analyses. I modify F. Sigaut’s tripartite breakdown of function to correspond to the documented psychological concepts to propose a pragmatic and theoretically palatable system for classifying lithics based on a combination of the causality of the tool’s physical properties, the embeddedness in the user’s biological system, and the recorded use history. These can be objectively defined and gleaned from archaeological specimens through the study of wear traces. I discuss the necessary methodological improvements upon which such a classification method is predicated and consider the necessary investments that must be made to realize its potential.