Portal de Periódicos da CAPES

Contextualizing Knowledge

2020; Duke University Press; Volume: 129; Issue: 2 Linguagem: Inglês

10.1215/00318108-8018615

1558-1470

Geoff Pynn,

Ethics in medical practice

Philosophical innovations have a predictable life cycle. They emerge in seminars and conference talks, spread among the cognoscenti, propagate in journals and anthologies, and finally, if they survive long enough, enter the realm of the book. Epistemic contextualism is in its golden years. In addition to Ichikawa's book, contextualist monographs by Keith DeRose (two, in fact), Michael Blome-Tillmann, and Peter Baumann have appeared over the last ten years. To stand out in this field, you need a twist. Ichikawa's is that contextualism and knowledge first epistemology go well together. This is a surprising theme, and Contextualizing Knowledge is full of adventuresome ideas. I'm puzzled about the basic agenda at the heart of Ichikawa's project. I'll try to explain why below. But the book makes many fruitful points worth pondering, and presents a dazzling array of overlapping lines of creative argumentation. I recommend it to anyone interested in acquiring a deeper acquaintance with some of the major concerns of contemporary analytic epistemology.Contextualizing Knowledge covers a sprawling range of topics: knowledge, counterfactuals, evidence, justification, rational action, assertion, and belief, as well as a variety of debates in philosophy of language. A loose framework of interlocking ideas and “moves” connects everything into a more-or-less systematic whole. Each discussion is deeply engaged with current debates and recent literature, and hence impossible to summarize in a short review. I'll focus on chapters 3 and 6. A quick summary of the others: Chapter 2 defends an account of the semantics of counterfactuals that, together with Ichikawa's theory of knowledge, implies that “sensitivity” and “safety” (in the technical senses those terms have acquired in the cottage industry devoted to modal accounts of knowledge) are both required for knowing. Chapter 4 gives a contextualist development of the claim that justification is potential knowledge. Chapter 5 argues that your reasons for acting are all of what you know, again with a contextualist twist. Chapter 7 presents a contextualist account of belief, and defends a version of the provocative idea that you should believe only what you know. If you have a professional interest in any of these topics, the relevant chapter is well worth your careful attention.Ichikawa's contextualism is based on David Lewis's (1996). He says you know p just in case you believe it on the basis of some evidence E, where all of the E-cases are p-cases. So far, this isn't contextualism. But like Lewis, Ichikawa intends his proposal to have “a certain kind of metasemantic generality” (19). Since quantifier domains vary with context, the domain of E-cases will vary with context, too. He prescinds from the “ambitious” task of articulating a set of rules for which possibilities get lassoed in by the “all,” offering instead a sophisticated discussion of epistemic standards. He gives a novel, though, to my mind, unconvincing argument for contextualism having to do with some odd conditionals.1 But Ichikawa's main task is not to argue for contextualism. Rather, he wants to use contextualism to “help the knowledge first project to avoid counterintuitive consequences” (8).The knowledge first project is not easy to define. It was born, everyone agrees, with Timothy Williamson's Knowledge and Its Limits (2000). Williamson proposed turning the theory of knowledge on its head: instead of explaining the concept of knowledge in terms of belief, justification, and other more technical notions, he would use the concept of knowledge to explain everything else. Ichikawa echoes Williamson when he says that “knowledge is in some sense fundamental, and can be used to illuminate other states of epistemological interest” (4). But unlike Williamson, Ichikawa does not focus on the concept of knowledge. His version of knowledge first epistemology says that knowledge itself is “a more fundamental feature of reality” than other epistemic phenomena, “metaphysically prior” to them in the same sense in which green is prior to grue, and electrons to laptop computers (4–5).Two things about this agenda are puzzling to me.First, I can't see how to square the fundamentality thesis with Ichikawa's theory of knowledge. Isn't giving a necessary and sufficient condition for when S knows that p precisely what knowledge-firsters think you shouldn't try to do, because you can't succeed? Williamson unequivocally rejected the idea that knowledge is a “complex hybrid crying out for analysis into its internal and external components” (2000: 5).2 But Ichikawa's entire project seems centered on an attempt to provide just that. Ichikawa says that he is not trying to provide an analysis of knowledge. And yet so much of the book's argumentation depends upon the Lewisian definition of knowledge in terms of belief and evidence. A hard-nosed Williamsonian might wonder whether the fly hasn't quite been shewn out of the bottle. Perhaps Ichikawa's response would be to remind us that his claim is only that knowledge is “relatively” fundamental. But this seems like a very capacious view. Not many would disagree that there is something less fundamental than knowledge. Of course, slavish adherence to Williamson's agenda is not a requirement of truth-seeking. But I do wonder whether it's fair to treat a project that revolves around a traditional-looking analysis of knowledge as a member of the knowledge first family.Second, the contextualist framework complicates the metaphysical priority claim substantially. Ichikawa says that knowledge is a relatively fundamental state. But the core thesis of contextualism is that a subject may truly be said to “know” in one context and not to “know” in another, all without a change in any of their states. So if it refers to a state at all, “knows” must refer to different states in different contexts. Is the idea that each of those different states is fundamental? Or just a subset of them? Or something else? Ichikawa never addresses this issue head-on, but it seems to me crucial for understanding what the metaphysical priority claim is supposed to amount to.Chapter 3 defends a contextualist version of Williamson's claim that knowledge and evidence are equivalent. Ichikawa observes that you might think that equating knowledge and evidence makes his theory of knowledge vacuous, since in chapter 1 he defined knowledge in terms of evidence. I was initially expecting a discussion of coherentism here—an epistemic light dawning slowly over the whole, a rising tide of evidential support lifting all beliefs, that sort of thing. Instead, he says that his view “invites a kind of foundationalism” (89). A kind of basic evidence, it turns out, holds up everything we know. He proposes, in another nod to Williamson, that this basic evidence consists of “world-involving factive mental states” such as that picked out by I see that I have hands.The chapter closes with one of the bigger surprises in the book. Ichikawa says that your basic evidence counts as “knowledge” in every context. Accordingly, he calls it “super-knowledge.” Since, he thinks, you super-know the contents of your world-involving factive mental states, you're in a position to “know” whatever worldly truths those contents entail, even in the most skeptic-friendly contexts you can conjure up. For example, unless they can get you to abandon your belief that you're not a bodiless brain in a vat, a skeptic can never make it false for you to say you “know” that. This is a notable departure from the standard contextualist approach to radical skeptical arguments. Contextualism was first advertised in the epistemology literature as a way to assuage the anxieties of radical skepticism. Ichikawa suggests that we turn to invariantist strategies, such as Jennifer Nagel's account in terms of loss of confidence, for therapy instead (Nagel 2010).Chapter 6 concerns assertion. The claim that knowledge is the constitutive norm of assertion is perhaps Williamson's most famous knowledge first idea. The most pressing objection to the knowledge account is that it implausibly treats all false assertions as improper since nothing false can be known. Ichikawa's response doesn't appeal to contextualism about “knows”; instead, he replaces the knowledge norm with something else. He follows Robert Stalnaker (1974) in modeling contexts as sets of possibilities compatible with the presuppositions of the conversation's participants. Assertions aim to add their contents to the conversation's presuppositions. Ichikawa introduces what he calls the “knowledge relation,” which you stand in to a proposition, relative to a context, when your evidence eliminates the conversationally open possibilities inconsistent with it. When you're making false presuppositions, you can stand in the knowledge relation to falsehoods. Ichikawa's “incremental norm of assertion” says you may assert only those propositions to which you stand in the knowledge relation. Hence, the incremental norm allows some false assertions.Consider an example: We are standing before a table that is white but lit by red lights. Since the table looks colored, and we don't know about the funny lights, we falsely presuppose that it isn't white. You're color blind, so you ask me whether it is red or green. I answer, “It's red.” My assertion is false, hence unknown, and so (rather harshly) forbidden by the knowledge norm. But, given our false presupposition, the only open alternative is that it's green, and my evidence eliminates that possibility. So I stand in the knowledge relation to what I've asserted, relative to our context. The incremental norm thus permits my assertion.I think the incremental norm lets you assert just about anything, provided your interlocutors are gullible enough. Want to get away with telling someone that the earth is flat? Just find some folks willing to join you in presupposing that the only other possibility is that we are living on the bumpy flesh of a massive cosmic toad. You stand in the knowledge relation to the proposition that the earth is flat relative to that bizarre context. Perhaps this isn't a problem: I don't know what you should and shouldn't say to such people. Nonetheless, it's a notable feature of the norm, and it makes it significantly more permissive than most competitors on the market. Moreover, to reiterate: the knowledge relation isn't knowledge. It's not even something you can refer to as “knowledge.” The incremental norm is not an example of putting knowledge to work in explaining other things.So I don't think that Ichikawa really places knowledge first. Knowledge is certainly discussed quite a bit. And most of the principles he arrives at start out as straightforwardly knowledge first ideas. But with one exception (his account of reasons for acting), each of the principles he endorses bottoms out in something other than knowledge: sets of epistemically relevant possibilities; basic evidence and cognitive processing; doxastic states and the notion of rejection of a possibility. Ichikawa confronts this issue in the closing pages of the book, and reiterates that the knowledge first claim he wishes to defend is only one of relative fundamentality. This seems to me a Pyrrhic victory for the book's larger agenda. For Williamson and those excited by his project, knowledge is not just relatively fundamental; it is epistemologically basic. It is one of the two “central relations between mind and world” (the other being action) (Williamson 2000: 1). Ichikawa could not endorse this claim. But this is only to challenge the knowledge first credentials of his many proposals, not their truth. Taken on their own terms, they deserve careful attention from epistemologists.Thanks to Julianne Chung, John Hill, Joe Glover, Mack Sullivan, and Peter van Elswyk for helpful discussion of Ichikawa's book as well as to Ichikawa and my co-commentators on his book (Kurt Sylvan and Alex Worsnip) at the 2018 Pacific Division meeting of the American Philosophical Association.