Xindi Ye
University of Hong Kong
My goal in this paper is to show that there is a dilemma of assessment for conceptual engineering.
Conceptual engineering is the activity of assessing and improving our words and concepts (Cappelen 2018). There are many standards (epistemic, metaphysical, ethical, prudential) we can use to assess concepts. Which standards should we choose? Following McPherson and Plunkett (2021), I call this the Standards Question.
There are two ways to respond to the Standards Question:
Invariant. The ultimate standard governing the goodness of a concept is fixed across all contexts of use.
Context-sensitive. The ultimate standard governing the goodness of a concept changes according to facts about a context of use.
Simion (2018) endorses Invariant. She thinks that all concepts have the central function of giving us knowledge about the world. Revisions that compromise a concept’s central epistemic function are never permissible. Nado (2019) endorses Context-sensitive. She denies that there are central, invariant conceptual standards. We should prioritise different standards depending on our goals and purposes.
But neither Invariant nor Context-sensitive is fully satisfactory. Consider Haslanger’s (2000) ameliorative redefinition of ‘woman’. Suppose the redefinition is politically good—it proves pivotal for the pursuit of gender justice, but epistemically bad—it obscures underlying biological reality, poses obstacles to inquiry. If Simion (2018) is right, then we should reject Haslanger’s ‘woman’. But not only would we give up huge political gains, we also contradict a basic tenet of conceptual engineering: that we can engage in conceptual revision precisely for the sake of effecting certain (social, political) worldly changes. If Nado (2019) is right, then we should accept Haslanger’s ‘woman’ just in case the political gains outweigh epistemic losses. But how would we distinguish Haslanger’s project with that of, say, re-defining ‘stray dog’ to mean something like minor god, assuming this saves millions of canine lives each year? If we think that the ‘stray dog’ case borders on conceptual perversion, then the question is how, on Nado’s account, we should distinguish cases of legitimate conceptual revision from cases of conceptual perversion. Herein arises a dilemma for conceptual assessment.
Chair: Nicolas Cuevas-Alvear
Time: September 12th, 10:40 – 11:10
Location: SR 1.005 (online)
SOPhiA 2026 