Knowledge as a thick concept: explaining why the Gettier problem arises

Abstract

The Gettier problem has stymied epistemologists. After Edmund Gettier (1963) proved the insufficiency of the tripartite analysis of knowledge, his paper was followed by a barrage of reformulated analyses, most of which were returned with potential counterexamples. 1 A succession of reformulations and counterexamples continued for over thirty years. However, within the past decade the general impetus to solve the Gettier problem has significantly receded (with only a few exceptions 2). But the abatement of this project is not due to any consensus over one particular analysis-quite the opposite. Despite the incredible effort invested in the project, no uncontroversial analysis has emerged. Indeed, some even believe the number of inadequate analyses indicates that no correct one is forthcoming (e.g. Williamson 2000, 30). Whether or not the Gettier problem is resolvable, we still must face an important question: Why does this problem arise in the first place? So far, philosophers have seen the Gettier problem as either a problem peculiar to the concept of knowledge, 3 or else an instance of a general problem about conceptual analysis. 4 But I would like to steer a middle course. I think the problem arises because knowledge is a particular kind of concept known as a thick evaluative concept, and a Gettierlike problem is just what we should expect from attempts at analyzing a concept of this sort. 5 These claims unfold as follows.

Tracing thick and thin concepts through corpora

“…Chappell (2013), who generally questions the existence of thin concepts, similarly claims that “if there are any thin concepts in aesthetics, perhaps beautiful is one of them” (Chappell, 2013, p. 187). Others have been more skeptical defending the thickness of aesthetic and epistemic terms (Kyle, 2013; Roberts, 2018). Our results do provide some support for authors like Zangwill and Chappell, at least in suggesting that most of the nonmoral thick terms we investigated are thinner than moral thick terms.…”

“…In order to illustrate the lack of consensus and underline the importance of answering questions like those listed above, let us highlight just a few recent controversies. For instance, no consensus exists on whether legal concepts like  and  (see, e.g., Enoch & Toh, 2013;Topham, 2016), epistemic concepts like  and  (see, e.g., Kotzee & Wanderer, 2008;Kyle, 2013;Roberts, 2018;Väyrynen, 2021), emotional concepts like  and  (see, e.g., Díaz & Reuter, 2020;Phillips et al, 2017), concepts linked to the domain of purity like  (see, e.g., Curry et al, 2019); Haidt, 2007), and other concepts like  (Sytsma et al, 2019),  (Knobe, 2003), and   (Napolitano & Reuter, 2021) that play a central role in contemporary debates in philosophy and psychology are evaluative concepts. There is also no consensus on whether and (if so) how many thick concepts demonstrate variability with respect to their evaluative component.…”