Unpacking conceptual function as tool function

Abstract It is increasingly argued that conceptual engineering should be sensitive to the functions that representational devices serve for their users. Less clear, however, is how to interpret conceptual function . This paper addresses this question through what I call the ‘analogy-based strategy’, focusing on linguistic conceptual engineering that targets referential expressions. The analogy-based strategy builds on the observation that referential expressions are treated in conceptual engineering as a subset of human tools. Accordingly, it aims to develop an interpretation of conceptual function that takes seriously the analogy between conceptual functions and other tool functions, and draws on this analogy to reveal new theoretical roles that conceptual function can play in conceptual engineering. After introducing four desiderata for such an interpretation and explaining their significance, I critically assess two candidate interpretations of conceptual function that follow dominant philosophical approaches to function: the proper-function interpretation and the system-function interpretation. I show that both fail to meet key desiderata. As an alternative, I develop a novel motivation-based interpretation, which understands conceptual function in terms of those actual or potential effects of an expression that are central to why a user group is, often implicitly, motivated to retain it in their conceptual repertoire. I argue that this interpretation provides an attractive notion of conceptual function that meets all four desiderata and proves useful for certain kinds of conceptual assessment.