Disjunctive luminosity

Received: 10 July 2020 Revised: 20 January 2021 Accepted: 22 January 2021 DOI: 10.1002/tht3.486 ORIGINAL ARTICLE Disjunctive luminosity Drew Johnson Department of Philosophy, University of Connecticut, Storrs, Connecticut, USA Abstract Williamson's influential anti-luminosity argument aims Correspondence Drew Johnson, Department of Philosophy, University of Connecticut, Storrs, CT 06269. Email: drew.johnson@uconn.edu to show that our own mental states are not “luminous,” and that we are thus “cognitively homeless.” Among other things, this argument represents a significant challenge to the idea that we enjoy basic self-knowledge of our own occurrent mental states. In this paper, I summarize Williamson's anti-luminosity argument, and discuss the role that the notion of “epistemic basis” plays in it. I argue that the anti-luminosity argument relies upon a particular version of the basis-relative safety condition on knowledge. This commitment is significant because basic self-knowledge seemingly lacks any kind of distinct epistemic basis, such as inference, observation, testimony, etc., despite representing a genuine kind of knowledge of contingent matters of fact. I consider a disjunctivist account (due to Bar-On and Johnson), according to which true basic self-beliefs indeed lack an epistemic basis in any kind of epistemic method (such as inference), yet are still epistemically grounded in the mental states they concern. I argue that this account of self-knowledge is compatible with standard understandings of the basis relative safety condition on knowledge, but rejects the particular version required by the anti-luminosity argument. KEYWORDS anti-luminosity, disjunctivism, neo-expressivism, safety, selfknowledge, Williamson © 2021 The Thought Trust and Wiley Periodicals LLC. Thought: A Journal of Philosophy. 2021;1–9. wileyonlinelibrary.com/journal/tht3 1 JOHNSON 2 1 | INTRODUCTION Williamson's influential anti-luminosity argument (2000) aims to show that our mental states are not “luminous” to us, and that we are thus “cognitively homeless.” Among other things, this argument represents a significant challenge to the idea that we enjoy privileged basic selfknowledge—that is, the idea that our ordinary, unreflective beliefs about our current mental states are especially secure. If our own mental states are not luminous to us—if we are not generally in a privileged position to know we are in a mental state M, when we are—then it seems we cannot be said to have privileged basic self-knowledge in the first place. In the next section, I briefly summarize Williamson's anti-luminosity argument, and discuss the role that the notion of “epistemic basis” plays in it. Then, in section 3, I (i) discuss some desiderata on a satisfactory account of basic self-knowledge, (ii) present a schematic epistemological disjunctivist view of self-knowledge, and (iii) briefly indicate how the view meets the desiderata. Section 4 is devoted to showing how a straightforward response to the antiluminosity argument is made available by this disjunctivist approach to self-knowledge. 2 | THE ANTI-LUMINOSITY ARGUMENT According to Williamson, a condition is luminous just in case one is in a position to know that the condition obtains, when it does (2000, p. 96). The main candidates for conditions that might be luminous are present phenomenal mental states, such as pain, or coldness, etc. To set up the anti-luminosity argument, Williamson presents the “cold morning” case, summarized here (2000, pp. 96–97): Cold Morning: One morning you feel freezing cold at dawn. As the day goes on, you slowly feel warmer, until you feel hot by noon. Suppose that these changes are very gradual, such that you are not aware of any change in your feelings of cold over one millisecond. Suppose also that throughout the morning, you are continuously attending to how cold you feel. Now, we shall assume toward a reductio that feeling cold is a luminous condition. The basic anti-luminosity argument is, then, as follows (the presentation here largely follows Berker (2008) and Srinivasan (2015)): 1. If S feels cold at time ti, then S is in a position to know that S feels cold at ti (Assumed for reduction). 2. If S is in a position to know that S feels cold at ti, then S is feeling cold at ti + 1 (Premise). 3. At t0, S feels cold (Premise). 4. At tn, S does not feel cold (Premise). 5. At tn, S feels cold (From 1, 2, 3). Regarding the transition from premises 1, 2, and 3–5: the idea is that this follows from repeated applications of modus ponens for increasing values of i. Premise 4 and 5 contradict each other. By hypothesis, one feels cold at dawn and warm at noon; thus, premise 3 and 4. So we must reject either 1 or 2. Understood in this way, the main substance of the anti-luminosity argument resides in showing why we should reject 1, rather than 2. JOHNSON 3 What, other than intuition, could make premise 2 so plausible that it should be retained, rather than premise 1? One possibility, raised by Wong (2008), is that (i) if one knows that one is cold at ti, then one is cold at ti and (ii) if one is cold at ti, then one is cold at ti + 1. But (ii) is a soritical premise, and so we should avoid this possibility if we want to explain the superficial plausibility of premise 2 (as Srinivasan, 2015 points out). The alternative support for premise 2 that Srinivasan (with others) discusses is that the premise is based on some kind of safety condition on knowledge—a reading that finds support in Williamson's original argument. As Williamson presents it, if one knows one feels cold at ti, one's belief must be “reliably based” (2000, p. 97), where supposedly one's belief that one is cold at ti could not be reliably based if one did not also feel cold at ti + 1 a millisecond later. This is supposed to be so because one would be almost equally confident that one feels cold at ti + 1 as one did at ti, since one's confidence at ti + 1 has “a very similar” basis as one's confidence at ti. (Since one gradually warms up throughout the morning, the basis for one's belief at ti will apparently be very similar to the basis for one's belief at ti + 1, because the cold-feelings at each instant are similar—but more on this point later.) So, it initially seems that if one feels cold at ti but not at ti + 1 (or if one feels cold to slightly different degrees), one's belief that one feels cold at ti (to degree x) could not be “reliably based,” for there is a very nearby possibility in which one believes that one is cold (to degree x) when one is not (or one is cold to some degree y: y ≠ x).1 In other words, premise 2 appears to fall out of a safety condition, since having a safe belief requires that one's belief could not easily have been false. In order for a safety condition to be plausible, it must refer to the basis on which the belief is formed in the actual world. Suppose I form the belief that it is raining in the actual world on the basis of my vision, and there is a nearby possible world where I falsely form a belief that it is raining on the basis of testimony. If safety is not basis-relative, the result is that I do not know that it is raining in the actual world. But forming the true belief that it is raining by looking out the window to check is clearly knowledge-conducive in ordinary circumstances. Thus, I propose that we understand the safety condition supporting premise 2 of the anti-luminosity argument to at least include a condition about the basis for belief, as follows (see Pritchard, 2009 for a defense): Basis-relative safety: If one knows that condition C obtains on basis B, then one could not have easily falsely believed that condition C obtains on basis B. It seems to me that basis-relative safety is more plausible than plain safety, and Williamson has no good cause to resist formulating the anti-luminosity argument in terms of basis-relative safety. Indeed, it seems that basis-relativity plays a role in Williamson's own formulation of the argument, since he makes appeal to the reliable basis for one's belief that one feels cold, and he points out that the belief that one feels cold at ti and the belief that one feels cold at ti + 1 will have a “very similar basis” (97; 99; 101). This principle is silent on cases where S would easily have falsely believed that C obtains on some other basis, B*. In order for the anti-luminosity argument to succeed, then, it must be that a belief that C obtains on basis B fails to be safe if one could easily have falsely believed that C obtains on basis B0 in a phenomenally similar case. That is, premise 2 of the argument requires something like the following: Phenomenally similar basis-relative safety: If one knows that condition C obtains on basis B, then one could not have easily falsely believed that condition C obtains on basis B, or on a basis B0 in a phenomenally similar case (so as to be subjectively indistinguishable from the actual case). 4 JOHNSON While I grant the plausibility of the basis-relative safety principle, the plausibility of its phenomenal cousin is not so clear, especially in the case of self-knowledge. In assessing these matters, it will be important for us to clarify what the relevant basis could be in the cold morning case. While it may be uncontroversial that beliefs about ordinary empirical matters are generally formed on some distinct epistemic basis (such as inference, testimony, perception, etc.), it is far less clear what the basis is, and even whether there is a basis, for our ordinary beliefs about our current psychological states. This is of course crucial in the context of the anti-luminosity argument, since that argument targets basic self-knowledge. I will argue that when the basis for basic self-belief (or rather, the lack thereof) is properly identified, we can see that the basisrelative safety principle does not support premise 2 of the anti-luminosity argument, and that the phenomenal cousin of this principle does not hold. Here are two ways in which we might understand “epistemic basis.”2 On the first—call it basis1—the epistemic basis for a belief just is whatever epistemic method the subject employs in arriving at the belief, such as inference, testimony, etc. Employing an epistemic method amounts to putting in some cognitive effort; one must apply the method in forming an opinion on the relevant matter. I leave it open whether a subject must have reflective awareness of the basis1 for her belief to be epistemically justified in holding it. What is essential here is that where a belief has a basis1, that basis comprises the way that one knows. However, we should also recognize the possibility that some sorts of belief may be epistemically warranted even in the absence of a basis1. For example, the category of epistemic entitlement represents a way in which a belief can be warranted even though the believer need not employ any epistemic method in order to earn such warrant. As a category of warranted though baseless1 belief, separate from the notion epistemic entitlement (as typically construed),3 we can consider the notion of grounded belief (Bar-On & Johnson, 2019, pp. 324–326): Belief that is warranted directly by the state of affairs that renders it true, where there then is no distinct epistemic method that must be employed as an intermediary between that state of affairs and one's belief about it. Call such a category of epistemic grounding a basis2 for belief.4 In the next section, I consider an account of self-knowledge that explains the apparent baselessness of such knowledge as owing to the fact that basic self-belief lacks a basis1, and that instead basic self-beliefs have a basis2. I argue the basis for self-belief proposed by this account does not support the phenomenally similar basis-relative safety condition needed for the antiluminosity argument. 3 | E P I S T E M O L O G I C A L D I S J U N C T I V I S M AB O U T SELF-KNOWLEDGE In this section, I briefly present some desiderata on an adequate theory of self-knowledge, and then I present a disjunctivist view of self-knowledge and briefly indicate how it purports to meet the desiderata. I lack the space to fully compare this view to alternatives; my main goal here is to establish the viability of such a view, in order to discuss its implications for the antiluminosity argument. Basic self-knowledge seems to differ in substantial respects from knowledge of ordinary empirical matters of fact, though it is a matter for further discussion exactly what the surface features of basic self-knowledge are or how they are best construed. To a first approximation, here are some of the surface appearances of basic self-knowledge relevant to the present paper5: JOHNSON 5 Epistemic immediacy: beliefs about one's own current mental states appear to be epistemically baseless (at least in the sense of basis1). They do not, on their surface, appear to be based in any distinct epistemic method, including inference, perception, testimony, etc. Substantive self-knowledge: Basic self-beliefs about one's current mental states generally amount to an especially secure though fallible genuine knowledge of those states. These features capture what is plausible in the idea that knowledge of our own mental states has a special epistemic status, but without assuming overly strong claims about that status. For instance, it is not maintained that subjects are infallible when it comes to knowledge of their own mental states, and indeed it does seem that on occasion, in unusual circumstances, we can have false basic self-beliefs.6 Part of the challenge of developing an adequate account of self-knowledge amounts to accommodating each of the features above: for it seems prima facie puzzling how basic selfknowledge could be at once substantive while also being epistemically immediate. Here, the earlier discussion of epistemic basis becomes relevant. Bar-On's neo-expressivist theory of self-knowledge (2004, inter alia)—as well as some versions of constitutivism (e.g., Boyle, 2009; Shoemaker, 1996)—advance an approach to selfknowledge according to which basic self-beliefs lack a basis1. While such views are well-poised to explain the immediacy of self-belief, one might wonder how this non-epistemic approach could also explain substantive self-knowledge. Bar-On and Johnson (2019), elaborating on the neo-expressivist view, add that basic self-knowledge has what I have labeled a basis2, in terms of what Bar-On and Johnson call “epistemic grounding,” as described above (2019, pp. 324–326). In avowing (either in speech or in thought) one's mental state M, one's M itself both makes true and warrants the self-belief that one is in M, through being the rational cause of the avowal (Bar-On, 2004, p. 249ff). As intuitive support for the idea that one's mental state M is a rational cause for one's avowing, consider: if someone were—oddly—to ask why you believe that you are in pain, and this is not meant as a question about what you think caused the pain, the most natural answer, if any is, would seem to be: “Well, because I am in pain!,” or “I just am” (Bar-On & Johnson, 2019, p. 323). One's first order mental state itself rationalizes the avowing of that state. To elaborate: on Bar-On's neo-expressivist account, avowals are expressive acts in which subjects give voice to their mental states using linguistic vehicles that semantically self-ascribe the very states expressed, as in: “I want to go for a run.” According to Bar-On, avowing one's M not only expresses one's M, but also expresses one's belief that one is in M (2004, pp. 307–310). On this view, it is my desire to go for a run that provides reason for my avowing “I want to go for a run,” as well as for my belief that I want to go for a run. Since my desire is itself my reason for making the avowal, and so also for the occurrent self-belief, we can say that my self-belief is epistemically grounded in the mental state it concerns, despite not being arrived at through the deployment of any distinct epistemic method. The neo-expressivist view sketched above represents an instance of epistemological disjunctivism about self-knowledge. Epistemological disjunctivism—ordinarily presented as a view about perceptual knowledge—is the view that ordinary veridical beliefs of a certain kind are warranted by reflectively accessible and factive reasons, whereas corresponding false beliefs lack such warrant, even though the “good” and the “bad” case may be subjectively indistinguishable 6 JOHNSON (see Pritchard, 2012 for discussion and defense in the case of perceptual knowledge). The neoexpressivist view described above is disjunctivist because the warrant for the basic self-belief that one is in M in the “good” case is said to be the very mental state M itself; this reason is reflectively accessible, as it is one's reason for avowing, and it is factive, since one only has this reason if the self-belief is true. (It must be emphasized that what makes the view disjunctivist is not specifically the distinction between epistemic ground and epistemic method. Rather, the view is disjunctivist because epistemic grounding for basic self-belief is said to be reflectively accessible and factive.) The resulting account explains the immediacy of avowals, insofar the view denies that basic self-beliefs have any basis1. Additionally, since the self-beliefs expressed in avowing are epistemically warranted by the mental states they concern, the view can also make sense of substantive selfknowledge. Moreover, as Bar-On and Johnson (2019) argue, the view avoids a major objection to epistemological disjunctivism about perception. The objection is that it remains mysterious why a subject's reflectively accessible warrant for a perceptual belief could change depending only on factors external to the functioning of her perceptual system, which is the epistemic “intermediary” between external perceptible states of affairs and her perceptual beliefs. That is: How could the epistemic significance of the outputs of one's perceptual system vary when the inputs (e.g., light impinging on the retina) remain the same, having only different distal causes (Burge, 2005)? By contrast, in the case of self-belief, there is no epistemic “intermediary” at all between the relevant belief and the state of affairs it concerns and which makes it true, and which could then be common to both good and bad cases.7 When it comes to perceptual beliefs, one's belief can fail to be true just in virtue of an uncooperative environment (e.g., a perfectly convincing hologram); but it seems that when a basic self-belief is false, this is always due to some psychological malfunction internal to the subject, rather than a “brute error” due to an uncooperative environment.8 Given the neo-expressivist idea that basic true self-beliefs are positively epistemically warranted just by their being directly rationally caused by the states they concern, the possibility of “counterfeit” subjectively indistinguishable first-order mental states are not epistemically relevant to warrant in the good case.9 Even if there are such counterfeits, a capacity to rule out their possibility in a given case, or to be able to subjectively distinguish the counterfeit from the genuine article, is not part of the structure of epistemic warrant for true basic self-belief in the first place. So even the modal proximity of counterfeits does not undermine the warrant for basic true self-belief. This is exactly the point of contrast between self-belief and perceptual belief that, according to Bar-On and Johnson (2019), makes the neo-expressivist disjunctivist view of self-knowledge more plausible that disjunctivism about perceptual knowledge. 4 | DISJUNCTI VE L U MI N OS I T Y I now consider the implications of the epistemological disjunctivist view of self-knowledge just discussed for the anti-luminosity argument. At the end of section 2, I noted that “basis,” as it appears in the basis-relative safety principle (and its phenomenal-similarity cousin) supporting premise 2 of the anti-luminosity argument, admits of two readings, basis1 and basis2. The disjunctivist picture of self-knowledge just discussed explains the warrant for true self-belief in terms of basis2. The result of this is that true basic self-beliefs enjoy basis2-relative safety. The basis2 for a true basic self-belief is also the truth-maker for the belief. Thus, there can be no modally near cases (or any cases, for that matter) where one forms a false self-belief on the same basis2 as the corresponding true basic self-belief. Necessarily, the basis for the false self-belief (if it has one) JOHNSON 7 will be a different mental state from the basis for the corresponding true self-belief in the good case, and will likely result from some psychological malfunction/confusion. Neo-expressivism also provides reason to reject the phenomenal-similarity version of safety, at least for self-knowledge. For, presumably, what would make the possibility of a “counterfeit” mental state, in a case phenomenally similar to the “good” case, epistemically relevant to one's veridical self-belief in the actual world would be that the epistemic credentials of one's true self-belief rested on a discriminative capacity that could then be “duped” by a counterfeit. But neo-expressivism maintains that our self-knowledge is epistemically immediate in such a way as to not require application of a discriminative capacity to know one's own mind. This immediacy, in turn, is explained by the expressive character of avowals; the point that an act of avowing is an exercise of an expressive capacity to directly give voice to one's present mental states in a semantically articulate way. Thus, given the disjunctivist approach to self-knowledge sketched above, the phenomenally similar basis-relative safety principle does not hold. The whole idea of epistemological disjunctivism is to deny the assumption that one's reflectively accessible warrant for belief in a veridical case could only be as good as the corresponding warrant in a subjectively indistinguishable case where one's corresponding belief is false (the idea being rejected here is what McDowell, 1982 calls the “highest common factor” thesis). So, simply from the fact that one might easily falsely believe that condition C obtains in a non-actual case phenomenally similar to the actual case where condition C does obtain, it does not follow that one cannot know that C obtains, when it does. Applied to the cold morning case, we have it that there is no point in the morning that one has an unsafe belief that one is feeling cold. For even if there is a point where one believes one is feeling cold when one is not, this false self-belief cannot share its basis with the true selfbelief formed just before it at ti-1. Significantly, this approach to the argument can grant the anti-luminosity assumption that there may be cases of false self-belief indistinguishable from a case of true self-belief - including the case where one continues to believe one is feeling cold just after the moment in the morning when one has ceased to feel cold.10 It goes beyond the scope of the present paper to consider the implications of my conclusions for Williamson's overall project. However, it may seem that Williamson himself should be sympathetic to the disjunctivist approach here, given that he would likely resist the “highest-common-factor” reasoning, the rejection of which motivates disjunctivism, in which case the present paper may reveal tensions internal to Williamson's project. It is worth noting as well that if the Williamsonian idea that knowledge is itself a kind of factive mental state is combined with a disjunctivist account of self-knowledge according to which our mental states are luminous to us, we seem to be led to endorsing a version of the controversial KK thesis, according to which when one knows that P, one can know that one knows. (Indeed, this consideration may be part of the reason Williamson finds it important to make the anti-luminosity argument—see Williamson, 2000, pp. 11–12). In sum: if it is right to deny, at least for basic self-knowledge, that the warrant for belief in a good case can only be as good as the warrant for belief in a subjectively indistinguishable bad case, premise 2 in the anti-luminosity argument is false, and the argument fails to show none of our states are luminous. Ordinarily, when one can express one's mental state through avowing it, one can know that one is in that state. That is, we are at least disjunctively luminous. A C K N O WL E D G M E N T S I would like to thank Dorit Bar-On and Crispin Wright for their encouragement and discussion on this topic, and for helpful comments on earlier versions of this paper. I would also like to think anonymous reviewers at Thought for their comments; the reviewer's comments were JOHNSON 8 particularly helpful for clarifying how the neo-expressivist account of the structure of warrant for true self-belief bears on the phenomenally similar basis-relative safety condition discussed in the paper. ORCID Drew Johnson https://orcid.org/0000-0001-7507-7879 E N D N O T ES 1 In order for the argument to be plausible, we must assume that throughout the morning, one's judgments are always judgments of the form “I am feeling cold to degree x,” rather than simply “I am feeling cold” (Williamson 2000: 98–99). For there likely will be no point where one transitions simply from feeling cold to feeling not cold tout court. But it is far less clear that we have privileged basic self-knowledge of states of feeling cold to degree x, as opposed to self-knowledge of feeling cold simpliciter. One might thus concede that states of feeling cold to degree x are not luminous, yet contend that this does nothing to undermine the point that feeling cold simpliciter is a luminous condition. Still, a likely response will be that there must be some point in the morning where one transitions from feeling simply cold to feeling almost-cold. So even feeling cold simpliciter may fail to be luminous if the argument goes through. Thanks to Crispin Wright for raising this issue. 2 The distinction here is drawn from the discussion in Bar-On and Johnson (2019: 324–326), but with slightly different terminology. 3 See Altschul (2011); Burge (1993); Dretske (2000); Wright (2004) for discussion. 4 Although “basis2” is intended as a term of art, there is some risk that the term may mislead, at least insofar as “basis” suggests a deliberate application of a recognitional or discriminative capacity. Still, I think the term is apt, as it marks a source of positive reason to think true. Additionally, this choice of terminology makes clear the relevance of epistemic grounding to basis-relative safety. Bar-On and Johnson (2019) emphasis the baselessness of self-knowledge—but there is no substantive disagreement here, since Bar-On and Johnson intend “basis” to always involve a distinct epistemic method, which is what I have labeled as basis1. 5 I here follow Bar-On's (in progress) articulation of the distinctive surface features of basic self-knowledge. See also Bar-On, 2004: Chapter 1. I have omitted other significant surface features sometimes going under the labels “Authority” and “First-Person Privilege.” 6 For instance, consider a case where one is sitting in a dentist's chair, and one mistakenly believes that one is already in intense pain owing to one's anticipation, even though the drill is only approaching and has not yet contacted one's tooth. (This case is described and discussed in Bar-On, 2004, p. 322). 7 A similar view can be found as well in some comments of Wright's (2015) (thanks to Crispin Wright for pointing this out to me). According to Wright, when it comes to instances of perceptual knowledge, one always could have acquired the knowledge instead only through indirect non-perceptual grounds (e.g., testimony), whereas a piece of phenomenal self-knowledge could not have been based only on some indirect ground. Phenomenal self-knowledge, Wright suggests, is necessarily knowledge of a state of awareness, so there is no sense to be made of coming to know of such a state through purely indirect means. So, like the neo-expressivist, it would turn out that there is neither need nor room for indirect epistemic mediation between one's phenomenal states and one's knowledge of them. 8 One may worry that such psychological malfunctions should count as aspects of an uncooperative environment since an agent will not be in a position to tell when such malfunctioning occurs. However, what matters to the purported advantage of disjunctivism about self-knowledge over its perceptual analog is whether there is a relevant epistemic intermediary common to both good and bad cases, not whether malfunctions are owing to internal versus external factors. I set this issue aside here, since it is not essential to the argument of this paper to show that disjunctivism about self-knowledge is more plausible than disjunctivism about perception. Thanks to an anonymous reviewer for raising this issue. JOHNSON 9 10 9 For discussion of the epistemic irrelevance of possible “counterfeit” mental states, see Bar-On (2012). See also Doyle (2019) for discussion of this point as it relates to several disjunctivist accounts of self-knowledge. The possibility of granting this assumption is apparently not available to certain constitutivist responses to the anti-luminosity argument (see, e.g., Coliva, 2016; Wright, 1998; Shoemaker, 1996; Zimmerman, 2006), since such constitutivist theories are committed to holding that feeling cold and one's beliefs about whether one feels cold are not independent of each other. As Crispin Wright has pointed out, however, it is unclear whether we should want, in the first place, to say that cases where one does not feel cold might be subjectively indistinguishable from cases where one does feel cold. Perhaps that is right; but again, at least it is an option for the disjunctivist proposal here to grant the assumption. R EF E RE N C E S Altschul, J. (2011). Epistemic entitlement. The Internet Encyclopedia of Philosophy. www.iep.utm.edu/ep-en/ Bar-On, D. (2004). Speaking my mind: Expression and self-knowledge. Oxford University Press. Bar-On, D. (2012). Externalism and skepticism: Recognition, expression, and self- knowledge. In A. Coliva (Ed.), The self and self-knowledge (pp. 189–211). Oxford University Press. Bar-On, D., & Johnson, D. (2019). Epistemological disjunctivism: Perception, expression, and self-knowledge. In C. Doyle, J. Milburn, & D. Pritchard (Eds.), New issues in epistemological disjunctivism (pp. 317–344). Routledge. Berker, S. (2008). Luminosity regained. Philosophers' Imprint, 8(2), 1–22. www.philosophersimprint.org/008002/ Boyle, M. (2009). Two kinds of self-knowledge. Philosophy and Phenomenological Research, 78(1), 133–164. https://doi.org/10.1111/j.1933-1592.2008.00235.x Burge, T. (1993). Content preservation. The Philosophical Review, 102(4), 457–488. Burge, T. (2005). Disjunctivism and perceptual psychology. Philosophical Topics, 33(1), 1–78. https://doi.org/10. 5840/philtopics20053311 Coliva, A. (2016). The varieties of self-knowledge. Palgrave Macmillan UK. Doyle, C. (2019). Ringers for belief. In C. Doyle, J. Milburn, & D. Pritchard (Eds.), New issues in epistemological disjunctivism (pp. 345–365). Routledge. Dretske, F. (2000). Entitlement: Epistemic rights without epistemic duties? Philosophy and Phenomenological Research, 60(3), 591–606. https://doi.org/10.2307/2653817 McDowell, J. (1982). Criteria, defeasibility, and knowledge. Proceedings of the British Academy, 68, 455–479. Pritchard, D. (2009). Safety-based epistemology. Journal of Philosophical Research, 34, 33–45. Pritchard, D. (2012). Epistemological disjunctivism. Oxford University Press. Shoemaker, S. (1996). The first-person perspective and other essays. Cambridge University Press. Srinivasan, A. (2015). Are We Luminous? Philosophy and Phenomenological Research, 90(2), 294–319. https://doi. org/10.1111/phpr.12067 Williamson, T. (2000). Knowledge and its limits. Oxford University Press. Wong, W. (2008). What Williamson's anti-luminosity argument really is. Pacific Philosophical Quarterly, 89(4), 536–543. https://doi.org/10.1111/j.1468-0114.2008.00335.x. Wright, C. (1998). Self-knowledge: The Wittgensteinian legacy. In C. Wright, B. Smith, & C. Macdonald (Eds.), Knowing our own minds (pp. 13–46). Oxford University Press. Wright, C. (2004). Warrant for nothing (and foundations for free)? Aristotelian Society Supplementary, 78(1), 167–212. Wright, C. (2015). Self-knowledge: The reality of privileged access. In S. Goldberg (Ed.), Externalism and selfknowledge (pp. 49–74). Cambridge University Press. Zimmerman, A. (2006). Basic self-knowledge: Answering Peacocke's criticisms of Constitutivism. Philosophical Studies, 128, 337–379. https://doi.org/10.1007/s11098-004-7797-y How to cite this article: Johnson D. Disjunctive luminosity. Thought: A Journal of Philosophy. 2021;1–9. https://doi.org/10.1002/tht3.486