The Skeptic and the Veridicalist: On the Difference Between Knowing What There Is and Knowing What Things Are 1009462296, 9781009462297, 1009243322, 9781009243322, 9781009243308

Table of contents :
Cover
Title page
Copyright page
The Skeptic and the Veridicalist
Contents
1 Introduction
1.1 A Standard Skeptical Argument
1.2 Topics to Set Aside
2 Veridicalism
2.1 Three Paths to Veridicalism
2.1.1 Chalmers’ Structuralist Veridicalism
2.1.2 Valberg’s Phenomenological Veridicalism
2.1.3 Davidson’s Interpretation Veridicalism
2.1.4 A General Definition of Veridicalism
2.2 Distinction from Idealism and Externalism
2.3 Veridicalism and What Things Are
3 Problems for the Veridicalist Antiskeptical Strategy
3.1 No Further Possibilities Are Ruled Out
3.2 “Tabbies” versus Tables
3.3 No Further Advantage Posited by Veridicalism
4 Objections and Further Implications
4.1 Knowing That This Is Reality
4.2 Is Ignorance about Whatsk Things Are Merely Scientific
Humility?
4.3 Is Ignorance about Whatsk Things Are Merely Metaphysical
Humility?
4.4 The External World versus Other Minds
4.5 It Doesn’t Matter Whatsk Things Are
4.5.1 Pragmatically
4.5.2 Other Values
4.6 A Historical Worry: Have I Changed the Topic?
5 General Implications for Skepticism and Antiskeptical Strategies
5.1 Beyond Veridicalism to Other Rejections of (2)
5.2 All Successful Strategies Must Deal with (1)
5.3 The Metaphysical Question and the Standard Formulation
References
Acknowledgments

Citation preview

The Skeptic and the Veridicalist On the Difference Between Knowing What There Is and Knowing What Things Are

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Cover image: The Image Bank / Getty Images

Series Editor Stephen Hetherington University of New South Wales, Sydney

Epistemology

The Skeptic and the Veridicalist

About the Series This Elements series seeks to cover all aspects of a rapidly evolving field, including emerging and evolving topics such as: fallibilism; knowinghow; selfknowledge; knowledge of morality; knowledge and injustice; formal epistemology; knowledge and religion; scientific knowledge; collective epistemology; applied epistemology; virtue epistemology; wisdom. The series demonstrates the liveliness and diversity of the field, while also pointing to new areas of investigation.

AVNUR

This Element explores the nature and formulation of skepticism about the external world by considering an important antiskeptical strategy, “veridicalism.” According to veridicalism, even if you are in a skeptical scenario, your beliefs about the existence of ordinary objects are still true. For example, even if you are in a global simulation, things such as tables exist as simulated objects. Therefore, your ignorance of whether you are in such a scenario does not negate your knowledge that there are tables. This strategy fails because it raises an equally troubling skepticism about what such objects are: is the table you now see a simulated object? That this is equally troubling suggests that the core skeptical problem is about what the causes of our experiences are, regardless of whether they count as ordinary objects like tables. This motivates a reconsideration of the standard formulation of the skeptical argument, and undermines some other anti-skeptical strategies as well.

Yuval Avnur

ISSN 2398-0567 (online) ISSN 2514-3832 (print)

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Elements in Epistemology edited by

Stephen Hetherington University of New South Wales, Sydney

THE SKEPTIC AND THE VERIDICALIST

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

On the Difference Between Knowing What There Is and Knowing What Things Are Yuval Avnur Scripps College

Shaftesbury Road, Cambridge CB2 8EA, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467 Cambridge University Press is part of Cambridge University Press & Assessment, a department of the University of Cambridge. We share the University’s mission to contribute to society through the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781009462297 DOI: 10.1017/9781009243308 © Yuval Avnur 2023 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press & Assessment. When citing this work, please include a reference to the DOI 10.1017/9781009243308. First published 2023 https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

A catalogue record for this publication is available from the British Library. ISBN 978-1-009-46229-7 Hardback ISBN 978-1-009-24332-2 Paperback ISSN 2398-0567 (online) ISSN 2514-3832 (print) Cambridge University Press & Assessment has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain,accurate or appropriate.

The Skeptic and the Veridicalist On the Difference Between Knowing What There Is and Knowing What Things Are Elements in Epistemology DOI: 10.1017/9781009243308 First published online: December 2023

Yuval Avnur Scripps College

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Author for correspondence: Yuval Avnur, [email protected] Abstract: This Element explores the nature and formulation of skepticism about the external world by considering an important antiskeptical strategy, “veridicalism.” According to veridicalism, even if you are in a skeptical scenario, your beliefs about the existence of ordinary objects are still true. For example, even if you are in a global simulation, things such as tables exist as simulated objects. Therefore, your ignorance of whether you are in such a scenario does not negate your knowledge that there are tables. This strategy fails because it raises an equally troubling skepticism about what such objects are: is the table you now see a simulated object? That this is equally troubling suggests that the core skeptical problem is about what the causes of our experiences are, regardless of whether they count as ordinary objects like tables. This motivates a reconsideration of the standard formulation of the skeptical argument, and undermines some other anti-skeptical strategies as well. Keywords: epistemology, skepticism, metaphysics, external world, ontology © Yuval Avnur 2023 ISBNs: 9781009462297 (HB), 9781009243322 (PB), 9781009243308 (OC) ISSNs: 2398-0567 (online), 2514-3832 (print)

Contents

1 Introduction

1

2 Veridicalism

8

3 Problems for the Veridicalist Antiskeptical Strategy

18

4 Objections and Further Implications

28

5 General Implications for Skepticism and Antiskeptical Strategies

45

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

References

50

The Skeptic and the Veridicalist

1

Although the whole of this life were said to be nothing but a dream, and the visible world nothing but a phantasm, I should call this dream or phantasm real enough, if, using reason well, we were never deceived by it. Leibniz (1923)

1 Introduction Skepticism about the external world has been formulated in many ways through the centuries. It has been taken to concern not only knowledge, but justified belief, rational degrees of confidence, doubt, and certainty; it has taken the form of a theory, a challenge, a paradox, a way of life, and an invitation to doubt. These days, skepticism is usually associated with the question,

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Can you know that external, ordinary objects such as tables exist? This concerns what there is, or what things there are. Sitting at your table, you look down and see its surface, and feel its pressure under your hand. Is the table really there, or is it part of a vast computer simulation, or perhaps an idea in the mind of Descartes’ demon? If you conclude that you don’t know whether some such skeptical scenario is playing out, then you are supposed to conclude that, therefore, you don’t know that there is a table there. This is nothing special about tables, so the conclusion generalizes: you don’t know whether any external objects exist. This standard formulation ignores the question of what things are, or the ultimate nature of the things whose existence is in question. The nature of things such as tables is reasonably taken to be a question of metaphysics or science. In fact, skepticism calls into question whether the metaphysics and science of ordinary objects even have a known subject matter! The possibility that you are in a simulation or a victim of Descartes’ demon is an epistemological problem about what there is, and metaphysical questions which presuppose the existence of objects such as tables depend on a solution to that problem. But notice that this standard formulation of skepticism is not entirely metaphysically innocent. There is a metaphysical conception of external world objects already at play, and required for the skeptical reasoning to get off the ground, to the effect that tables are not the kinds of things that could exist in a simulation, or in the mind of a demon. And accordingly, one might try to stop skepticism in its metaphysical tracks by rejecting this conception of objects. Berkeley famously proposed that the external world is ultimately ideal, or made up of objects that are themselves ideas about which there can be no serious or even coherent skeptical worry. This would solve the problem of skepticism, perhaps. But this sort of idealism seems too extreme to most of us today, perhaps even more extreme than skepticism itself. Instead, in this Element we will explore a different metaphysical, antiskeptical strategy, one that rejects some standard metaphysical assumptions about the external world, but stops short of idealism.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

2

Epistemology

Setting aside Berkeley’s idealism until Section 2, consider this alternative metaphysical view: If this is all a vast simulation, then tables still exist. In that case, tables turn out to be simulated or virtual objects. If this is all one big dream, then tables still exist. In that case, this is a table “in the dream,” and tables turn out to be ideas rather than mind-independent matter. What tables are depends on which scenario you’re in. So, goes the alternative approach, there’s no problem. Look down again at your table. When you realize that you don’t know whether or not you are in a simulation or a victim of Descartes’ demon, you should not conclude from this that therefore you don’t know whether there is a table. If this is all a simulation or a demon scenario, then the table turns out to be a virtual object or an idea in the mind of Descartes’ demon. That’s just what things are in those scenarios. If you don’t know whether you are in some such skeptical scenario, then although you know what there is – there is a table – you don’t know what things are – what is this table, a computation in a simulation, an idea in the demon’s mind? But that is no skeptical catastrophe. You don’t know whether things are ultimately simulations or ideas in a demon’s mind, but this is no different from perfectly acceptable metaphysical humility about the ultimate nature of things. We should already accept that the best metaphysics or science will not necessarily tell us everything about what things are, and that’s all the skeptical possibilities show. This alternative way of thinking swaps skeptical ignorance of what there is for skeptical ignorance about what things are. It is at least as old as Leibniz, but following its most recent proponent, Chalmers (2018), I call it veridicalism. The key point of this Element is that veridicalist skepticism about what things are is just as epistemically disastrous as standard skepticism about what there is. It isn’t mere metaphysical humility. As we will see, veridicalism might address specific formulations of skepticism, but it does not solve the problem posed by the skeptical argument. In other words, the metaphysics of ordinary objects affects the formulation, but not the epistemic significance, of the skeptical conclusion. And this will reveal something important about the problem of skepticism. A satisfactory solution to the skeptical problem requires dealing with the part of the formulation that doesn’t depend on one’s metaphysical outlook. It takes getting clear about what part of the skeptical problem does depend on our metaphysics to see this. An entirely reasonable reaction at this point is: But how could this be? The alternative, veridicalist metaphysics vindicates the claims that you know that there are tables, that you can put cups on tables, that tables are made of wood, and so on. (You merely lack knowledge of whether all of those things are ultimately simulations, or ideas.) How could this fail to solve the problem of skepticism about the external world? It shows that you know a lot! Consider that

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

3

a small child, or our ancestor in the ancient world perhaps, knows a lot about what there is. This knowledge is significant to their lives, and they know how to act in light of this knowledge. And yet, we can imagine, they either have no beliefs about what things are beyond how they appear (in the case of the child), or else they have a bunch of false beliefs about this (in the case of our ancestor, who perhaps thought that all things were ultimately made of water, or fire). Whatever it is that one worries about when one worries about the skeptical problem – that we know nothing about the world – surely these worries don’t apply to the cases we are imagining, as described! Yet, all these people lack, it seems, is knowledge about what things are. So, how could veridicalism, which posits that we are roughly in the same position, since we know what there is but not what things are, possibly fail to satisfy us? It shows that we know at least as much as the child or the ancestor, and their situation does not seem worthy of philosophical panic. The bulk of the arguments in this Element aim to show that this attitude, though initially appealing, is mistaken. The child’s and the ancestor’s ignorance about what things are is not the same as skeptical ignorance about what things are. As we will see, the former is a matter of immaturity (whether of the individual or of the science of their time). The latter is a matter of necessary, complete hopelessness. And, as we will see, it is equivalent in important ways to ignorance about what there is. It is true that not all kinds of ignorance about what things are constitutes an epistemic disaster. But a particular, skeptical sort of ignorance about what things are is different. And so, one can know a lot about what there is while knowing very little about the external world. Whether skeptical hypotheses raise doubts about what there is or what things are is, as veridicalism helpfully shows, a matter of metaphysics. But if this table could be anything from a simulation to a demonic idea, then the mere conclusion that you know that there is a table is still compatible with almost total ignorance about the external world. As I will suggest, this also serves to clarify what a skeptical scenario is, and its relation to other metaphysical hypotheses about the world. Skepticism isn’t really only about knowledge of what there is, rather it is standardly formulated that way because of a standard metaphysical picture. Here is an overview of what is to come. In the rest of this seciton, after I introduce the standard, contemporary skeptical argument, I describe and set aside some more familiar, nonveridicalist antiskeptical strategies. In Section 2, I introduce the veridicalist strategy and some of the different ways it has been pursued. This strategy avoids skepticism about what there is only by accepting skepticism about what things are. I then distinguish veridicalism from some other, externalist and idealist theories, and describe skepticism about what things are and how veridicalism is committed to it.

4

Epistemology

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Sections 3 and 4 are where most of the action (and novel argumentation) occurs. Section 3 presents arguments for thinking that skepticism about what things are leaves us with as little knowledge about the world as skepticism about what there is. One important consequence of this is that the veridicalist antiskeptical strategy fails – at least, if you were ever worried about standard skepticism about what there is, then your worry should not be assuaged by veridicalist skepticism about what things are. In Section 4 I consider some objections to my arguments: veridicalism at least grants us knowledge that the world we see around us is reality; skepticism about what things are is not a radical skepticism at all, but rather mere scientific or metaphysical humility about the ultimate nature of things; it doesn’t matter what things ultimately are pragmatically or for other values; the problem of skepticism is only about the existence of things, and other concerns about the nature of things are a change of topic. In reply to this last objection, I discuss some of the history of skepticism, from Sextus to Hume and Moore, to show that the exclusive focus on the existence of ordinary objects is a relatively recent phenomenon, and that skepticism about what things are has often been taken to be a concern. Finally, in Section 5 I describe some more general lessons that can be learned from the way in which veridicalism fails to solve the skeptical problem. One can know what there is without knowing much about the world. Some other antiskeptical strategies besides veridicalism fail for this reason, and furthermore the standard argument is inadequate as an expression of the skeptical problem. I also make some observations about what a successful antiskeptical strategy would have to look like.

1.1 A Standard Skeptical Argument To begin, consider one currently standard formulation of skepticism and the skeptical argument. Skepticism, as Barry Stroud’s (1984) seminal work formulates it, is the view “that no one knows anything about the world” (p. 1). If true, that would be epistemically devastating. But what exactly is knowledge “about” the world? Philosophers these days interpret this to be a matter of what there is: you don’t know that there are tables. How is this skeptical conclusion reached? This standard skeptical argument revolves around skeptical scenarios which purport to cast doubt on vast swaths of our beliefs. In his Meditations on First Philosophy, Descartes appeals to such scenarios, for example one in which you are being deceived by an all-powerful demon. So this standard skeptical argument is often referred to as the “Cartesian” skeptical argument. In Descartes’ demon scenario, the demon makes you have the experience you are currently having, so that it seems like there is an external world full of ordinary things such as tables

The Skeptic and the Veridicalist

5

and people. But, in fact, there are no such things – at least according to the standard description of the scenario. Let us focus on such global skeptical hypotheses, which on a standard metaphysics call all of our knowledge (or justified beliefs) about the external world into doubt.1 For examples, the hypotheses that your entire life has been like one long dream, that you are a brain in a vat being stimulated by a computer in an otherwise empty world, and that you are in the scenario depicted in the film The Matrix since birth are all global skeptical hypotheses. Let “sk” stand for your favorite such hypothesis:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

(1) (2) (3) (4)

I do not know that ~sk. If I do not know that ~sk, then I do not know that there are tables. I do not know that there are tables (from (1) and (2)) Generalizing from tables, I do not know that there are any ordinary things (hands, tables, sandwiches, etc.).2

(I will remind the reader what each of (1)–(4) is each time it is used in a new section.) This is one standard formulation of the argument, but a few qualifications are helpful here. First, arguably the dream version of sk raises a different set of issues, which I address in Section 2. Second, one might reasonably construe the “standard” skeptical argument as the argument for (1), leaving the existence of ordinary objects as a further, different matter. Set those arguments aside here, since our concern is the relation between global skeptical hypotheses and the existence of ordinary objects. Third, there has been some discussion in the literature about so-called underdetermination principles which generate (3) and (4) in an alternative way. Veridicalism can be understood to apply to those versions of the argument as well, but at least the latest versions, due to Chalmers and Valberg, seem to focus on (1) and (2), so I will follow them.3 Ultimately, I conclude that identifying the skeptical problem with (4) is not quite right, since some views on which 1

2 3

I briefly address one veridicalist’s (Chalmers’) view on nonglobal skeptical scenarios in my review of his book, Avnur (2023). This formulation closely follows Chalmers’ (2018, 1). One recent and helpful discussion of the relation between underdetermination principles and the sort of closure principles that are typically taken to motivate (1)–(4) is Pritchard (2016, especially ch. 2). The underdetermination principle states, roughly, that if two propositions are (known to be) incompatible, then in order to know either one of them one’s evidence must favor that one over the other. Briefly, the underdetermination version of the skeptical argument is: (1*) My evidence does not favor the proposition that there are tables over sk. (2*) In order to know that there are tables, my evidence must favor the proposition that there is a table over sk. (by the underdetermination principle) (3) I do not know that there are tables. (from (1*) and (2*)) So, (4). Generalizing from tables, I do not know that there are any ordinary things.

6

Epistemology

(4) is false still fail to avoid the sort of epistemic disaster that makes (4) problematic. It is not necessarily enough to show that we know what things there are.4 Why is (4) such an epistemic disaster? The ignorance (4) posits implies that your cognitive life, in relation to the world, is a joke: for all you know, you are thinking about the world and your place in it completely wrong. You don’t even know that there are tables, and the most you can know is that there are tableish sensory experiences, which is a nightmarish relation to have to reality. In other words, (4) makes explicit just how little you know about the world. Although I will mostly be setting this aside, external world skepticism is not all about tables, it’s also about people. As Stroud put it when contemplating the skeptical conclusion:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Other people, as I understand them, are not simply sensory experiences of mine; they too, if they exist, will therefore inhabit the unreachable world beyond my sensory experiences, along with the tables and chairs and other things about which I can know nothing. So at least with respect to what I can know I could not console myself with thoughts of a like-minded community of perceivers all working together and cheerfully making do with what a communal veil of perception provides. (Stroud 1984, 38)

You know nothing about things, or people, beyond your sensory experiences. Though the existence of other people is a central issue for those who worry about classic skeptical scenarios, as we will see in what follows, there is some question whether veridicalists mean to be addressing such scenarios at all. If they intend to address only those scenarios in which other people still exist (as in The Matrix), then they don’t address a significant part of what has been worrying philosophers about (1)–(4). I will leave this point aside, though, to see how veridicalism fares even with respect to a narrower set of skeptical hypotheses, and even setting the existence of other people aside.5

4

5

The veridicalist strategy here would presumably be to reject the claim that the proposition that there is a table is incompatible with sk, and so to maintain that one can know it without possessing the requisite evidence that favors it over sk. Accordingly, (2*) is not supported by the underdetermination principles after all, and can be rejected. Other strategies may go the route of rejecting the principle, or rejecting (1*). But the veridicalist, it seems, can accept these and still hold that (2*) is false. Much of what I say about veridicalist objections to (1)–(4) will apply to this version as well, and I note that at least two recent proponents of the veridicalist strategy, Valberg and Chalmers, explicitly target (2) rather than (2*). For the record, Stroud, in a different essay, seemed to agree, at least about the point that what things there are does not settle what sort of world we are in. See Stroud (1986, 263–264). See Helton (forthcoming).

The Skeptic and the Veridicalist

7

To be sure, some people, and even some philosophers, don’t find (4) so worrying at all. What I will argue in the next section is consistent with that: if one finds standard skepticism about what there is disturbing, then one’s worries should not be assuaged by veridicalism.

1.2 Topics to Set Aside Since the focus in this Element is the metaphysical underpinnings of the standard skeptical argument, and since there is a large variety of antiskeptical strategies and some overlap between them, it will be helpful to explicitly set aside strategies and views that, though independently interesting, can distract from our target. Recall the standard argument:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

(1) (2) (3) (4)

I do not know that ~sk. If I do not know that ~sk, then I do not know that there are tables. I do not know that there are tables. I do not know that there are any ordinary things.

Philosophers try to avoid (3) and (4) by rejecting (1) or (2). The most direct strategy is to reject (1): show that your information is sufficient to rule out sk. Denying (2) is indirect in the sense that it targets what seems to follow from our ignorance about sk, rather than sk itself. Though I will be focusing on a particular kind of indirect approach, and setting aside direct approaches, the direct approaches will come up again in a couple of contexts. First, there are well-known strategies, discussed in the next section and associated with Hilary Putnam’s “semantic externalism” and George Berkeley’s idealism, that superficially resemble veridicalism but are crucially different in various respects. As we will see in Section 2, Putnam-style externalism aims to show that we can rule out sk a priori, rather than showing that even if we don’t know whether sk, we could still know that there are tables. We will also see that Berkeley’s idealism is incompatible with various skeptical hypotheses (and posits a metaphysics of objects that is incompatible with veridicalism), and so also rules them out a priori. So these views reject (1), rather than (2). We will mostly set them aside. Second, such rejections of (1) will come up again in Section 5. As we will see, the only way to satisfactorily solve the skeptical problem about the external world, insofar as this problem arises from consideration of skeptical hypotheses such as sk, is to pursue the direct approach of rejecting (1). In other words, much of what follows can ultimately be taken as an argument for the direct strategy as superior to the indirect one.

Epistemology

8

So, our focus is on the indirect strategy, which accepts (1) but rejects (2). But even here there is plenty of variety, and plenty to set aside. One familiar indirect strategy is to deny (2) for epistemological reasons. Implicit in the justification for (2) is the idea that, since the existence of tables implies the falsity of sk (recall that sk is a global skeptical scenario), knowing that there are tables implies knowing that not-sk. But if that is the justification for (2), then one way to avoid commitment to (2) is to avoid commitment to such conditions for knowledge. Maybe it isn’t always the case that knowing something, p, and knowing that it entails something else, q, puts you in a position to deduce or otherwise know q. The considerations involved in this kind of strategy, often called “denying closure,” are all epistemological, since they involve what is required to know something. That is, this strategy does not challenge the idea that if there are tables, then sk is false. It does not challenge the underlying metaphysics of objects. So, for the most part, we will set these strategies aside. However, as with direct strategies, these epistemological indirect strategies will come up again at the end, though this time not as beneficiaries. As we will see, much of the trouble for veridicalism is also trouble for some epistemological indirect strategies as well.

2 Veridicalism

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

In this section I describe a few versions of the veridicalist antiskeptical strategy and explain its commitment to skepticism about what things are. They each deny (2): (2) If I do not know that ~sk, then I do not know that there are tables, on the basis that there can be tables even if sk is true, because what appear to be tables in any global skeptical scenario are tables. So, accordingly, we don’t need to know whether or not we are in a skeptical scenario in order to know that there are tables.6 Crucially, veridicalism is compatible with (1): (1) I do not know that ~sk. Part of the appeal of the view is that it need not bear the burden of explaining how we know ~sk. Veridicalism can seem far-fetched. How could something in the mind of a demon, or in a simulation, be a table? Tables are real things, we want to say, and while simulations themselves can be real, what they are simulations of are 6

Notice that this does not require rejecting closure for knowledge. For a different take on the relation of (2) to skeptical hypotheses, see Roush (2010) and Avnur, Brueckner, and Buford (2011).

The Skeptic and the Veridicalist

9

not, by virtue of being simulated, real. Let us now see how a few different philosophers motivate an alternative, veridicalist metaphysics.

2.1 Three Paths to Veridicalism Veridicalism has been around at least since Leibniz (1923) and, in the last century, Bertrand Russell (1927) and O. K. Bouwsma (1949).7 More recently, veridicalism has been defended in different ways by David Chalmers (2005; 2012, 431–440; 2018; 2022), J. J. Valberg (2007), and, on one prominent interpretation, Donald Davidson (1986). Here I will discuss the latter three.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

2.1.1 Chalmers’ Structuralist Veridicalism Chalmers has defended veridicalism in various works over many years (Chalmers 2005; 2012, 431–440; 2018; and 2022). He arrives at veridicalism through structuralism about physical properties.8 Roughly, structuralism about some property is the view that that property is individuated (or the term expressing the property defined) by its causal role (or by its relation to other terms in a theory) (2018, section 3). For instance, having mass is having a, or being the, property that plays the mass role (2018, 11). This also applies to other physical properties which things such as tables have, such as spatial properties. Details aside, the result is that being ignorant about sk does not itself imply that we are ignorant about whether there are tables. For, even if sk is true, there is something playing the role(s) associated with being a table, and that is a table.9 Presumably, on this view, if the table is a demonic idea (for example), then the fact that it has the causal role of a table amounts to the fact that it causes, in some sense of “cause” that we will grant the structuralist, other demonic ideas to occur as well as the relevant experiences in the mind of the victim. Oddly, when you knock on the wood of the table, on this view, you are knocking the demon’s mind, and when you taste the coffee that is in the cup on the table, you are tasting the demon’s mind. And if you are in the Matrix, you are tasting a computer chip in a certain state! Whether one is in a simulation or a demon victim will make some difference to the causal profile of the table. If the table is a demonic idea, then a demon 7 8

9

Bousma is discussed in Chalmers (2018, 627–629). Sklar (1985, 60–161) and Vogel (1990, 661) both consider, and reject, a sort of structuralist veridicalism view as well, as Chalmers points out. See Chalmers (2018, 24–27) for some important details, including whether mass is identified with the property that actually realizes the requisite causal role, or whether it is identified as the property of having a property that plays that role. Though this is an important distinction for Chalmers, it won’t play an important role here.

10

Epistemology

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

could, if it wanted, make the table disappear. Some (perhaps those more comfortable with theistic views according to which God constantly sustains all objects) might find this to be perfectly compatible with what they usually take the causal profile of tables to be, while others might find this to be inconsistent with how they assume tables behave. Presumably, though, all have tables in mind. So, in some respects, the causal profile of tables permits of some flexibility. Perhaps there is some range of different causal profiles that, if satisfied, is sufficient for tablehood. So, different “kinds” of tables will have different total causal profiles. But structuralist veridicalism holds that your ignorance about whether sk is true does not imply that you do not know that there are tables, because you know that something within that range exists. You still know that they have a certain mass, that they have four legs, that you can put cups on them, and so on; so (2) is false. On this view, sk is a metaphysical hypothesis about the ultimate nature of things like tables, and perhaps also about the specific causal profile of the table. If you are a demon victim, there is still a table here, but the table is ultimately a demonic idea that would vanish if the demon willed it. If you are in the Matrix or a brain in a vat (BIV), then the table is ultimately a computational state and could vanish according to the program, or by a glitch. And so on for other sk’s. There are metaphysical objections to structuralism, independent of its implications for skepticism, that I will set aside.10 What I will argue in what follows is that structuralism (and other versions of veridicalism, and some other denials of (2)), even if true, would not suffice to solve the global skeptical problem. 2.1.2 Valberg’s Phenomenological Veridicalism Valberg’s (2007) view is not intended to solve all relevant skeptical problems, but only the standard sort. One cannot do his nuanced phenomenological observations justice in summary form. But, roughly, he observes that the objects of experience are arrayed around you in a spatial expanse with you (the subject) at the center. He calls this array, or that within which things appear to you, your “personal horizon” (Johnston 2010, ch. 2, following Valberg in most details, posits a similar phenomenon, calling it your “arena of presence.”) According to 10

Vogel (2019) presents some, focusing on whether concepts such as distance and causation can be construed structurally. See Chalmers (2018, 18–19) for replies. But Vogel agrees with Chalmers that structuralism, if true, would solve the global skeptical problem (pp. 3–4), so on this point I differ from them both. Another sort of problem for structuralist veridicalism might be developed from Korman’s (2015, 39) argument that our ordinary object concepts are not so undemanding as to apply to whatever is out there in our environment.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

11

Valberg, if this is a dream, then the table you see in front of you is a table “in the dream,” and this is an extrinsic, rather than intrinsic fact about the table (2007, 30, 85, 93). As he puts it, the “this” in “this is all a dream” refers to your personal horizon, which in this scenario is a dream. You can know, in the ordinary way, that there is a table, then, without knowing whether this is a dream; you can know what the thing appearing in front of you is without knowing what that within which it appears is. “Whether THIS [your present personal horizon] is a dream does not bear, one way or the other, on whether this is a cup (say), an external object . . . dream skepticism is transcendent . . . It leaves my knowledge of this totality, whatever it is, untouched. It leaves untouched, then, precisely the knowledge that skepticism about the external world calls into question” (110–111). On this view, skepticism is “both vindicated and disarmed” (113). What is left is a “transcendent” skepticism about whether the world around you, in which things like tables appear, is a dream. But this does not bear on whether you know that there is a table, on Valberg’s view. Ultimately, my criticism of veridicalism is compatible with this latter claim about transcendent skepticism, but I will argue that Valberg’s treatment of the more standard (what he calls “immanent”) skepticism does not constitute an adequate solution to it. Valberg applies the same strategy to other global skeptical hypotheses, such as the BIV, but most of his remarks focus on dreams. So, a note about dream hypotheses is in order. One might think that, in one sense of “dream,” dream experiences are, by definition, nonveridical. Valberg, for various reasons, rejects this view, and holds that if you are dreaming that there is a table here, there is indeed a table, and that table is ultimately ideal, or mind-dependent, since it is a table “in a dream.” So, when we are considering, with Valberg, the dreaming scenario as a veridicalist would interpret it, we must keep in mind that this is a hypothesis about the metaphysical nature of the table: if you are dreaming this, then the table exists, and is a “dream table,” an item in the dream. But one could be a veridicalist who doesn’t accept this view of dreams. For this reason, from here on, except when discussing Valberg specifically, I avoid the dream hypothesis altogether and stick to demons, BIVs, and the Matrix. 2.1.3 Davidson’s Interpretation Veridicalism Davidson’s (2006) antiskeptical view rests on a view of interpretation rather than structuralism about properties or personal horizons. As with Chalmers’ structuralism, all the details won’t matter here, so a rough sketch will suffice. The objects of most of one’s beliefs are to be identified with their actual causes,

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

12

Epistemology

and thus must be interpreted to be largely correct (and the same applies to one’s assertions) (2006, 236). So, to say that your belief that there are tables is false is to misunderstand your belief. However, this applies only to one’s simplest beliefs – the simplest cases get “special weight” in this interpretation (2006, 237) – so that nonsimple, and more theoretical beliefs about the inner nature of tables are not immune to error due to interpretation considerations. And, at least on one plausible reading of Davidson, your belief that you are not a BIV (assuming you have such a belief) is not one of those simpler beliefs that are immune to error. Thus, this view is at least compatible with (1), since it does not guarantee that the truth-conditions of the belief that ~sk are met. And yet, your belief that there are tables must be, on this view, true on any correct interpretation, so (2) is false.11 Davidson was a veridicalist12 on this standard interpretation of his view.13 As with Chalmers’ structuralism, there have been some objections to Davidson’s view. For example, Craig (1990, 213), McDowell (1994, 17), and Brueckner (2010, 181) have argued that Davidson’s veridicalism swaps ignorance about whether the belief that there is a table is true for ignorance about what the belief that there is a table is about. Is it a belief about computer states, demonic ideas, or the sort of mind-independent physical object we typically assume it to be? Davidson, the objection goes, makes us ignorant about the contents of our beliefs. One might wave this worry away if one accepts, as many do, Putnam’s (1975, 227) saying that “meanings ain’t in the head.” Some ignorance about the contents of our beliefs may be the norm, rather than some skeptical catastrophe. But, more importantly for us in this context, not all versions of veridicalism directly or obviously involve ignorance about our belief contents, so such objections to Davidson would not apply to other versions of veridicalism. Recall Valberg’s veridicalism, for example, on which whether this is a dream table is an extrinsic feature, which, it seems, would not affect the content of the belief that there is a table at all. So, I will also set these objections aside.

11

12

13

Klein (1986) suggests that there is an important gap between securing the truth of one’s belief that there is a table, and knowledge that there is a table, so that (2) is not so easily secured. But, presumably, ordinary knowledge attributions are among those simpler, nonphilosophical beliefs that should be charitably interpreted. At any rate, I won’t pursue this, or other, objections to Davidson’s appeal to charity here. For interpretations of Davidson as a veridicalist – as accepting (1) (or at least not denying (1)) and rejecting (2) – see Brueckner (2010, 180), Button (2013, 142), Craig (1990), Klein (1986), and McDowell (1994, 17 fn 14). Though he doesn’t address Valberg, Chalmers rejects Davidson’s veridicalism, as well as Bowsma’s (1949) ordinary language appeal to “illusion,” which he considers to be a version of veridicalism.

The Skeptic and the Veridicalist

13

2.1.4 A General Definition of Veridicalism Here, then, is veridicalism as I will consider it. It has two parts: First: (2) is false because there are (knowable) things such as tables even if sk is true, and even if (1) is true; therefore, the argument from (1) & (2) to (3) (I do not know that there are tables) & (4) (I do not know that there are any ordinary things) is unsound. Second: (1) shows merely that we do not know something about the ordinary things (like tables) around us, for example whether they are simulations.

According to veridicalism, global skeptical hypotheses concern the ultimate nature of ordinary things, rather than the existence of ordinary things. So our ignorance about whether sk, as in (1), amounts to ignorance about the ultimate nature of ordinary things. Section 3 considers how problematic this ignorance is. But first, it will be helpful to distinguish veridicalism from a similar-seeming strategies which raise entirely different sets of issues, in order to avoid some common distractions.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

2.2 Distinction from Idealism and Externalism A distinctive feature of veridicalism is its acceptance of terms with which to formulate sk as an open possibility. For example, the table is mind-independent if the world is as you usually assume it to be, but mind-dependent if you are a victim of the demon. So, the meaning of “mind-independent” is neutral with respect to which scenario you are in. Chalmers (2005, 12, 18), calls such terms “semantically neutral” (he calls them “non-twin-earthable” terms in [2018]) and they include “computational,” “envatted,” “action,” and “stimulated.” With neutral terms, not only can we articulate skeptical worries about specific scenarios, but also distinguish, say, BIV from Matrix and demon scenarios (2005, section 8). Since all versions of veridicalism are compatible with (1), and countenance sk as an open, conceptual possibility, they all accept that there are some semantically neutral terms with which to specify some of sk’s details, and what exactly we might be getting wrong about the nature of ordinary objects.14 14

For more on why Chalmers thinks we can specify these details about sk, while some of our other terms, such as place names like “Tucson” are not neutral, see his (2005, section 9, note 1) and (2018, pp. 2–4) Valberg also thinks we possess neutral concepts enough to specify and distinguish different sk’s. For example, in contrast to the dreaming scenario: “the [BIV] hypothesis (properly understood) is . . . first, that there is a transcendent world, a world outside [this personal horizon]; second, that there is in the transcendent world a brain maintained and stimulated in a vat; and third, that the transcendent brain-in-a-vat setup is responsible for there being such a thing as [this horizon], with the world (including [me] and [this] brain) internal to it” (p. 118, fn 12). So, clearly ‘vat,’ ‘stimulated’, and ‘transcendent’ can be used neutrally. He also does not think that the

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

14

Epistemology

An alternative view rejects all neutral terms with which to formulate sk. If there are no neutral terms, then in any world – even one created by Descartes’ demon – everything counts as physical and mind-independent, and we cannot express what we are getting wrong about objects in a demon scenario. Such a strategy may appear similar to, or just a more extreme version of, veridicalism, but it rejects (1) instead of (2) – arguably, it must accept (2), since (2) would then have a false antecedent. It rejects (1) because, on this view, it is a priori that the table is mind-independent, not an idea, not a computation, and so on, so it is a priori that sk is false. One such strategy, which rejects (1), is Putnam’s (1981, 14) content externalism, on which concepts such as brain, vat, and even causation, are not semantically neutral (Putnam, 1994, 207). Rather, what they express is determined, roughly, by what you are actually in contact with – recall that, in contrast, on Davidson’s view this is true only of one’s simplest beliefs, not all of them. Putnam argued that, therefore, whatever scenario you are in, you are not a BIV, and thereby rejected (1). Putnam (1994, 287) and others who follow his antiskeptical strategy (for example Tim Button (2013, 117–48) and A. W. Moore (1996, 224–230; 2011, 51–52) also hold that you possess no concepts with which to describe what it is about the table that you are ignorant of due to (1). Your skeptical ignorance about the table is ineffable, because you lack the conceptual resources to specify it.15 So if you are somehow deceived, or are misconceiving the world, you cannot say how. It seems to follow that no real skeptical problem about the table due to sk remains.16 One problem with the Putnamian approach is that it seems to do nothing about skeptical scenarios in which you were recently envatted. This shortcoming it has in common with veridicalism, which applies only to global skeptical scenarios – the case of recent envattment and emerging from the vat will come up again later (Section 4.1). There is plenty of literature criticizing Putnam, and other such a priori rejections of (1), and I do not have much to add to that here.17 Instead, I grant, following veridicalism, that we have neutral concepts with which to express sk and our ignorance about tables due to sk, and that we can

15

16

17

dreaming and BIV scenarios are “pseudo-hypotheses,” as the Putnamian approach, described later in this section, suggests (121). Note that Button’s view on whether we possess the requisite concepts is more complex than I have sketched here; see chapter 19 of his book, on the “cosmopolitan” view, and the book’s Coda. Sk is merely a “bare formal” possibility, as Button (2013, 131–132) puts it. See also Button (2013, 147,48) on “metaphysical” skepticism, which holds “nothing more than a nebulous sense that the world is mysterious.” In contrast, veridicalists hold that there are specifiable and meaningful aspects of things that we fail to know about the table, due to (1). See Nagel (1986, 73), Wright (1992, 93), and Sundell (2016, 248).

The Skeptic and the Veridicalist

15

articulate the basic details about sk in its various versions. As we will see, my argument ultimately can be used to support such Putnamian approaches, since one upshot is that only views that reject (1) can solve the problem that the standard argument poses. A different, older view that might also be confused with veridicalism is Berkeley’s idealism. According to Berkeley, things such as tables exist, but they are ideal – ultimately, ideas in the perceivers’ minds (including God’s). Thus, their ultimate nature is discoverable: such objects are all mind-dependent. This differs from veridicalism in a number of ways. To begin with, veridicalism does not reject (1), and it is committed to the view that we do not know the ultimate nature of things. The ultimate nature of things depends on what scenario we are in, and we don’t know which scenario we are in. In contrast, Berkeley does reject (1) and rejects the view that we don’t know the ultimate nature of things. For example, Berkeley thinks we can rule out the BIV scenario, on which tables (according to veridicalism) are mind-independent, computersimulated objects. He thinks we know that ordinary things are ultimately ideal.18

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

2.3 Veridicalism and What Things Are In this section I clarify the idea that, according to veridicalism, though you know that there is a table, due to (1) you don’t know whether the table is a simulation, or an idea, and so on. This amounts to ignorance of what the table is, by which I mean what category to apply to it. There are different aspects of a thing, which, depending on the context, one might have in mind when thinking about what that thing is. For example, if you tell me I will inherit some furniture, but I don’t know whether it is a table or chair, that is one sort of ignorance about what it is. If you tell me it is a table, but I don’t know whether it is wooden or plastic, that is another sort of ignorance about what it is. If you tell me it is a table, but not whether it is an idea in some other being’s imagination, then that is yet another sort of ignorance about what it is, one that only veridicalism makes possible, and one that seems intuitively much more extreme, or at any rate unordinary. For present purposes, it is not necessary to come up with a complete account of “knowing what” – though see Dasgupta (2015) for one recent account that is compatible with my use of “knowing what.”19 Failing to know something about 18

19

See Chalmers (2022, 68–71) for a discussion of the differences between veridicalism and idealism. According to Dasgupta’s (2015, 468) account, “I ‘know what’ something is iff I can identify it by its nature.” Arguably, if you don’t know whether the table’s “inner nature” is computational, ideal, etc., you fail to know “its nature” in the sense Dasgupta develops, the “essence” of a thing.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

16

Epistemology

a table, or its nature, due to (1) constitutes one sort of ignorance about what the table is. So even if you know that whatever there is “here,” which causes your table experience, counts as a table, you do not know what kind of table it is – a demonic table, a simulated table, ad so on – according to (1). Since this ignorance derives from ignorance about sk, let us call it ignorance about “whatsk” something is. According to veridicalism, then, you don’t know whatsk the table is, and that is just to say that the aspect of it that you fail to know about is the aspect that you are ignorant about due to (1): whether it is an idea, a simulation, and so forth. If we accept (2), we can express ignorance about the table due to (1) as (3). If, instead, we reject (2) and accept veridicalism, then we are still committed to ignorance about tables, but the structure of this ignorance changes, from (3), to our failing to know whatsk the table is. The structure has changed from what there is (are there tables?) to whatsk things are (are tables simulations?). It is worth emphasizing the compatibility of ignorance about whatsk the table is with some other knowledge of what the table is, so as not to overstate veridicalism’s commitments. For example, you might still know that the table is not a sofa, is made of wood, has mass, and that you can put a cup on it, according to veridicalism. For, wood, mass, cups, and impenetrability are, on the structuralist version of veridicalism, analyzed in terms of their causal roles, which remain the same on sk. On Valberg’s version, wood, mass, cups, and impenetrability are all there in the dream. Whether they are in the dream is an extrinsic feature of them. And similarly, beliefs about the table’s having these properties must be interpreted to be true, on Davidson’s view. Thus, since the property of being made of wood, for example, is instantiated here whether or not sk is true, that aspect of the table is not included in whatsk it is. However, it is also important to note that this knowledge of what the table is does not extend very far. You still don’t know whether things like mass and wood are demonic ideas, for example. That is just more ignorance of whatsk things – in this case mass and wood – are. Note, too, that since tables, we are assuming, could have demonic causal profiles (as discussed earlier in this section), ignorance of whatsk the table is implies ignorance of whether, say, it would vanish if Descartes’ demon willed it. For those who don’t like this talk of “what” things are, similar remarks apply to the “nature” of a thing. Insofar as it is in the nature of a table that it is not a sofa, then according to veridicalism you can know the nature of the table. But there is another part of its nature that you cannot know due to your ignorance about sk, or (1): is the nature of the table that it is a demonic idea, a computational simulation?20 20

I avoid the use of “essence” because whether it applies depends on the specific version of veridicalism at issue. If a table is anything that has table-like causal roles (what Chalmers calls “role” structuralism), then sk does not specify part of the essence of a table, since the table could exist regardless of whether it is material, ideal, or computational. If instead tables are whatever in

The Skeptic and the Veridicalist

17

Consider the different skeptical claims at issue. Skepticism about what there is, I cannot know whether there is a table, because I might be a BIV or a demon victim,

is swapped by veridicalists for skepticism about whatsk things are: I cannot know whether this table is a BIV-table, a demon-table, and so on. (In other words, I cannot know whatsk the table is.)

Intuitively, it would be strange to find the first conclusion worrying while not finding the second conclusion worrying. To be sure, some people don’t find either one worrying or problematic. For such people, the standard skeptical argument, (1)–(4), does not constitute a philosophical problem, but perhaps (at most) a curiosity or trivial puzzle. For such people, there is no use for the veridicalist “solution” to the skeptical problem, because there is no problem. As I explain further later, my claim is ultimately a conditional one: if you find skepticism about what there is worrying, then you should be just as worried by skepticiam about whatsk things are. Intuitively, they seem to express the same basic ignorance about the world. The second conclusion about whatsk things are seems like a mere reformulation of, rather than an improvement over the first conclusion about what there is. This is because they are both implied by the following basic skepticism:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

I cannot know what is causing my table experience, a BIV machine, a demon, and so on.

Basic skepticism is implied by (1) on its own, regardless of whether we accept (2) or veridicalism. Once we notice that veridicalism does not avoid basic skepticism, it becomes hard to feel relieved by swapping the standard skepticism about what there is for veridicalist skepticism about whatsk things are. Basic skepticism is just what worried us about standard skepticism in the first place if standard skepticism was ever worrying at all: you do not know about the world beyond its sensory effects on you. This is the external world after all. It was only due to some metaphysical assumption about what tables could or couldn’t be that we expressed such worries as worries about what there is, rather than what things are. But that’s just how we expressed the same, basic worry about the external world. In the next section, Section 3, I vindicate this intuition with some arguments. the actual situation in fact has the table-like causal role (what Chalmers calls “realize” structuralism), and if in fact tables are ideal, then it is part of the essence of tables that they are ideal, because nonideal things can’t be tables. By focusing on nature instead of essence, we remain neutral on these two versions of verdicalism.

Epistemology

18

3 Problems for the Veridicalist Antiskeptical Strategy21 We have seen so far that the standard skeptical argument (1)–(4)

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

(1) (2) (3) (4)

I do not know that ~sk. If I do not know that ~sk, then I do not know that there are tables. I do not know that there are tables. I do not know that there are any ordinary things.

concludes with skepticism about what there is, and the veridicalist antiskeptical strategy leads to skepticism about whatsk things are. And I have suggested that these two skeptical conclusions are intuitively equally problematic because both are committed to basic skepticism: you don’t know what is causing your experiences, what is “beyond” your subjectively accessible mental states, or in the external world. In this section I will support this intuition with some arguments. Admittedly, it is a slippery matter what “problematic” in “equally problematic” amounts to, or what one should find worrying in the first place. So the first task is to get clearer on this. (1)–(4) is valid, and its premises look plausible, at least at first. The argument poses a problem if the conclusion, (4), is worrying or unacceptable from a philosophical perspective. As I have suggested, (4) is worrying because it implies that our relation to reality is a joke, because it implies that we have very little knowledge about the world around us. One might attempt to avoid this, and to assuage the worry, by showing that the argument (1)–(4) is unsound. But if such a refutation of (1)– (4) commits us to something just as worrying as (4), by also implying that our relation to reality is a joke, or that we have just as little knowledge about the world as (4) implied, then that refutation fails to solve the problem presented by (1)–(4). This is what I will be arguing. In other words, my conclusion here is a conditional one: If you found the standard skeptical conclusion (4) about what there is worrying, you should not be significantly less worried by the veridicalist skeptical conclusion about whatsk things are.

Earlier, I suggested that there is an intuitive equivalence between skepticism about what there is and skepticism about whatsk things are, so that the conditional is true. Here are some considerations that vindicate the intuition. 21

“In Avnur, Yuval (forthcoming) Veridicalism and Scepticism. Philosophical Quarterly. I raise some similar problems for a narrower class of veridicalist strategies, focusing on structuralism.”

The Skeptic and the Veridicalist

19

3.1 No Further Possibilities Are Ruled Out Let us say that a possibility is “ruled out” by one’s knowledge if it is incompatible with one’s knowledge. Veridicalism does not posit that we can rule out any more possibilities about reality, and about tables, than standard skepticism does. It is natural to understand the information one has as a function of which possibilities one can rule out, so it seems to follow that veridicalism does not imply that we have any more information about the world than standard skepticism does. Rather veridicalism reformulates that same amount of information that is posited by standard skepticism. One way to see this is to consider the table more closely. Set aside scenarios in which nothing is causing your experiences, so that they are randomly occurring. Veridicalists such as Chalmers, for example, have nothing to say about such scenarios either. Instead, focus on global scenarios in which something or other is causing your experiences, as in the classic skeptical hypotheses. Both the standard skeptic about what there is, and the veridicalist skeptic about whatsk things are, agree that we know the following disjunction:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

There is either a demonic table, or a computer-table, or a “non-sk”-table . . . and so on. The standard skeptic does not count knowledge of this existentially quantified disjunction as knowledge of the existence of a thing whose existence conditions are so vastly disjunctive, because she holds that only non-sk tables are tables. For veridicalists, as long as you know that the disjunction is satisfied in the actual situation, you know that there is a table. This is because whatever satisfies the disjunction in the actual situation is a table – and, relatedly, nothing else is a table.22 So, while you still don’t know which disjunct is satisfied, as long as you know that one of them is, you know that there is a table. The only difference, then, is that the veridicalist claims that you know that there is a table; both the veridicalist and the skeptic think that all you know is that something satisfies the disjunction, and that you cannot know which disjunct is true. But how can veridicalism posit more knowledge about the world than the standard 22

I am here presenting Chalmers’ (2018, 650) version of veridicalism, or “realizer structuralism,” according to which whatever actually realizes the relevant causal role is a table. It is arguable that, on Valberg and Davidson’s views, being a table is the even less specific property of being anything that satisfies the disjunction. That is, for example, for Valberg, a table in the dream is a table, but so is a table outside of a dream. There are advantages to either kind of veridicalism. Briefly, the advantage of the first kind is that a replica of an object (say, inside a holographic model) that plays the same causal role as the real object will not count as a real object. And that seems right. On the other hand, on that view, if you happen to be a BIV, then tables are computer states, and nothing else could be a table – not a demonic idea or even a regular, non-sk table. It seems odd, though, to say that a nonsimulated thing with the causal role of a table is not a table, since it seems intuitively to be a paradigm case of a table as we all (even we who are BIVs) conceive of tables.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

20

Epistemology

skeptic, if there is no possibility about the table, no disjunct, which can be ruled out according to the veridicalist but not the skeptic? The difference seems to be merely how we describe the possibilities that we cannot rule out.23 The veridicalist might respond by saying that on her view one possibility that can be ruled out is that there are no tables. This is denied by the skeptic who accepts (3) and (4). But this is a difference about how possibilities are described, not which possibilities can be ruled out. According to the veridicalist, the claim that we can rule out the possibility that there are no tables is just the claim that we can rule out the possibility that nothing plays the causal role associated with tables. However, the skeptic who accepts (4) can also rule that out. Something, either the demon, the BIV computer, the robot overlords in the Matrix, your dreaming brain, or something else, plays a role that we associate with tables. (Other skeptics might deny that there is even any causation, but here we are considering a skeptic who accepts (1)–(4) but accepts that something causes her experiences.) She knows this in just the same way that the veridicalist does: while granting (1) (I do not know that ~sk) and not knowing whatsk the thing causing the table experience is. So, the very same possibility can be ruled out by someone who accepts (4). The only difference is that the skeptic who accepts (4) does not say that anyone who can rule this out thereby knows that there are tables, because she does not hold the same, permissive view of what counts as a table. To be clear, the standard skeptic and the veridicalist agree that this possibility, which we can rule out, is one in which there are no tables, because playing the causal role associated with tables is necessary for being a table. They disagree only about whether whatever is actually playing that role counts as a table, even if it is an sk-entity. So, they disagree about how to describe the remaining possibilities. But there is no possibility that veridicalism says we can rule out, but which standard skepticism does not. Therefore, we possess the same information about the world according to veridicalism as we do according to standard skepticism, and this implies that veridicalism fails to solve the problem posed by (1)–(4). Another, related way to reach this conclusion is by appeal to skeptical neologisms, next.

3.2 “Tabbies” versus Tables By merely inventing some clever terms, the standard skeptic can know as much about the world as the veridicalist claims to know. Recall that we are setting aside scenarios (which veridicalists also do not address) in which nothing is causing 23

Note that this applies as well to an “underdetermination” skeptic, who holds that you do not know that there are hands because your evidence does not favor the hypothesis that there are hands over the hypothesis that you are in an sk scenario being stimulated in way that mimics seeing hands. This kind of skeptic also agrees with the veridicalist that there is either a BIV table, a demonic table, or a non-sk table (and so on).

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

21

your table experiences with any regularity. Both the veridicalist and the standard skeptic claim that we have knowledge that something is causing our table experiences and has a causal profile similar enough to what we usually think tables have. Even the skeptic knows that, if we are deceived by the demon, then our experiences are caused by the demon. The only difference is that the veridicalist claims that whatever this is counts as a table. But the skeptic could come up with a clever term, “tabbies,” which she stipulates to apply to the thing that actually causes her table experiences. She has thereby turned knowledge that something caused her table experience into knowledge that there are tabbies. So, she accepts (4), and yet now she also knows what things there are: tabbies. She knows as much about tabbies as the veridicalist claims to know about tables. Now our clever skeptic – who I call a skeptic because she accepts (1)–(4) – who knows that there are tabbies, may no longer qualify as a skeptic about the external world, because she claims to know what there is, namely tabbies (though she still holds that we don’t know whether there are tables). Has she thereby avoided the skeptical disaster implied by (4)? Compare what this clever skeptic, who accepts (4), claims to know and what the veridicalist claims to know. They both know that something causes their table experiences, but the veridicalist calls it a table, while the skeptic calls it a tabby. But knowledge that there is a table and knowledge that there is a tabby amount to the same information about the world. Metaphysically, there is no difference between tabbies and the veridicalist’s tables: they exist in all the same possible (global) scenarios, and regardless of whether sk is true. Veridicalists hold that the clever skeptic has a false theory of tables, as we ordinarily think of them, because tables are tabbies, even though this skeptic says that sk is incompatible with tables, but is compatible with tabbies. But that is a disagreement about what counts as a table; there is no substantial difference between their claims to knowledge about what is going on around them, given that that they both accept basic skepticism. The information they claim to have about their environment is the same, expressed differently. We can now see why the veridicalist antiskeptical strategy is not an adequate solution to the skeptical problem posed by (4). Granting veridicalism, (4) is false, and skepticism about what there is is avoided. (4) posed a problem by raising a worry about our knowledge about the world. But veridicalism’s solution to the problem posed by (4) is adequate only if the knowledge we have about the world according to veridicalism is substantially better than the knowledge we can have about the world according to (4). But someone who accepts (4), by merely stipulating new terms such as “‘tabbies,” can claim to have essentially the same knowledge as the veridicalist claims we have about the world, only using different terms to express it. So, the knowledge we have according to veridicalism is not substantially better than the knowledge we can have according to (4). It follows that veridicalism doesn’t adequately solve the skeptical problem posed by (4).

22

Epistemology

Another thing follows from this: the standard skeptical argument, (1)–(4), is not an adequate expression of the skeptical problem, since refuting the standard argument is insufficient for solving the problem. We will return to this important upshot later.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

3.3 No Further Advantage Posited by Veridicalism So far, I have argued that veridicalism does not posit more information (in the form of possibilities ruled out) than standard skepticism, and that standard skeptics posit all the same knowledge about the world, but in different terms, as the veridicalists. I claimed that this implies that veridicalism leaves us as badly off, with respect to our knowledge of the external world, as standard skepticism. Skepticism about what there is is no worse than skepticism about whatsk things are. But does veridicalism perhaps do something else that makes it a worthwhile response to standard skepticism? One might suggest that veridicalism vindicates our ordinary thinking about the world, since we ordinarily believe there are tables, not tabbies. If so, then at least veridicalism posits ordinary knowledge about the world, and this may be thought to be an improvement over our situation according to standard skepticism about what there is. In evaluating this claim, I set aside issues about the individuation of belief contents that are raised here – on some theories of content, a belief that there is a tabby has one and the same content as a belief that there is a table, so the objection is false. Still, there are other problems with this suggestion. To begin, interpreting this suggestion is a bit complicated because it is ambiguous between a dissolution and a solution to the skeptical problem. Consider the dissolution. On this interpretation, our ordinary belief that there are tables does not discriminate between sk-tables and non-sk-tables. So, it should not bother us that we don’t know whether there are any non-sk-tables. But that is all (1)–(4) could establish. Once it is clarified that, in (3) and (4), the skeptic means to deny us knowledge of non-sk objects in the external world, we should no longer be worried or disturbed. This is a dissolution of the skeptical problem: there is no problem, really, because nothing about skeptical conclusions like (4), or the fact that we cannot rule out any of the disjuncts discussed in Section 3.1, should worry or concern us.24 The reply to this dissolution is twofold. First, technically, what I have been arguing is entirely compatible with this because I aim to establish only 24

Davidson arguably intended his veridicalism as a dissolution of this kind. He did not see himself as answering the skeptic, so much as ridding us of the need to answer (2006, 238–241). Nevertheless, one might, and many have, appealed to Davidson’s view as a solution to skepticism.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

23

a conditional claim: if you find standard skepticism (1)–(4) worrying, then you shouldn’t take solace in veridicalism. This is compatible, of course, with the antecedent’s falsity. Second, and more substantively, it is implausible that our ordinary thinking about the world is vindicated by veridicalism. We ordinarily assume (or are disposed to believe) various things about tables that are not vindicated by veridicalism: that they are not demonic ideas, or that they are not items in a vast simulation brought about by robot overlords. This is clear from the fact that skeptical scenarios initially appear to contradict what we assume about the world, and surely you would be surprised if you somehow learned that your entire life, and the world that stimulates your senses, is actually a computer interface in a video game played by some being outside the universe you know. Some of our ordinary beliefs are vindicated (there are tables), and some of our ordinary beliefs (tables are not figments of some creature’s imagination) are not. The question remains whether saving only the first kind of belief is sufficient to undermine any worry we might have about the second – more on this later. To be sure, other assumptions about tables, for example that they are not mostly empty space, have turned out false, and learning this has not amounted to a full-blown skeptical crisis. I address this rebuttal, and the difference between learning something new about a table’s material nature and our ignorance of whether its nature is as according to sk, later, in Section 4. Moreover, it is difficult to square the preceding dissolution with the claim, established earlier, that veridicalism fails to posit any substantial knowledge or information about the world that (4) fails to posit. Whether to call the things causing our experiences “tables” rather than “tabbies” is a paradigmatically empty question: not a question about which scenario you are in, but how to describe that scenario.25 So, even if veridicalism saves some of our ordinary thinking about the world, it is unclear how this could matter. Veridicalism does not show that we know any more about the world, or about tables, than we do according to standard skepticism. This is due to the way (4) is avoided, namely by a very permissive account of what could count as a table. How could the way your knowledge is formulated or described make a difference to how worried 25

This notion is originally due to Parfit (1984, 235). He illustrates it in Parfit (2011, 435): When about to cross the channel, suppose you know that you will feel seasick, but you wonder whether being seasick counts as being in pain – this is a question about the extension of the term “pain,” not a question about how you feel. Here, you know what will happen, though you don’t know how to describe it. This is in contrast to a case in which, about to undergo some medical procedure, you don’t know whether you will be in pain. In the channel case, there aren’t two different possibilities: you know which possibility will hold, just not what to call it. As Parfit puts it, “It matters whether, while receiving the medical treatment, I shall be in pain. And it matters whether, while crossing the Channel, I shall be seasick. But it does not matter whether, in feeling seasick, I can be said to be in pain.”

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

24

Epistemology

you are about your knowledge of the world? Recall that veridicalism is still committed to basic skepticism, according to which you don’t know what causes your experiences. If that is what was worrying, then veridicalism cannot help. Turn now to the interpretation according to which veridicalism solves the problem raised by (4) because our ordinary beliefs, or more of them than according to (4) anyway, turn out to be knowledge after all. On this view, the number of ordinary beliefs that are true and constitute knowledge is higher according to veridicalism than it is according to (4), and this is enough of an epistemic difference to assuage the problem raised by (4). For example, you know that there is a table made of wood, and a cup on the table, and that the table and cup are in a room, and you even know what city the room is in, and so on. These are ordinary beliefs that constitute knowledge according to veridicalism, but not according to standard or even basic skepticism (though the latter leaves this open rather than denying it). Earlier, I argued that this is not epistemically significant, since it does not show that your information or knowledge about the world excludes any possibilities (it only changes how those possibilities are described), and that the very same items of knowledge can be known by a skeptic who accepts (4) and who also invents a clever neologism. Here, the objection is that the superior quantity of ordinary beliefs that count as knowledge is itself valuable, and this is an advantage that the veridicalist has over the standard skeptic. However, the comparative “score,” as I will refer to the matter of how many of our ordinary beliefs constitute knowledge, is, first, not so straightforward, and, second, not clearly significant. First, counting beliefs is notoriously a slippery thing, where one belief becomes a kaleidoscope of millions. For example, you believe that there is a table, a table in front of you, and also a table closer than ten feet from you, closer than twelve feet from you, and so on. How many beliefs to count there is not obvious. This muddies the veridicalist’s objection, because there is a similar, indefinite number of things you don’t know concerning the table, according to veridicalism, so that the net advantage to veridicalism over standard skepticism is difficult to discern. For one, you don’t know that the table is mind-independent, that it is not made to appear by a demon, that machines that have enslaved humanity didn’t cause it to exist, that they didn’t cause it to exist while running low on batteries, while being powered by an even number of solar panels, and so on. But surely you also ordinarily believe, or are disposed to believe, such things. Sure, these don’t often come up in ordinary life and are seldom said out loud, but neither does “there is a table in front of me.” We still ordinarily believe all of these things. If the latter counts as ordinary despite, as Wittgenstein famously pointed

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

25

out in On Certainty (with respect, most famously, to hands and physical objects), seldom coming up in conversation, then so do the former.26 It might be objected that beliefs such as that there is a table play a role in guiding one’s behavior and thinking, while the belief that the table is not, ultimately, a demonic idea does not. So, goes the objection, while the score itself might be indefinite, the importance of the beliefs vindicated by veridicalism is greater than those that remain unvindicated. But the belief that one’s life is not lived within a demon’s imagination or a vast simulation is surely a momentous one for one’s understanding of the world and one’s life in it, or one’s worldview (regardless of whether we consider this to be one belief or a conjunction of many), and this may affect one’s decisions and attitudes in the same way that, say, how seriously one takes one’s life can affect one’s decisions. It is worth pausing briefly here to consider this notion of understanding. There is some debate about how to analyze “understanding,” but it seems plausible that if you don’t know whatsk it is that is causing your experiences, then you don’t really understand, or you understand only to a relatively low degree, what is going on around you. This means, for example, that you don’t even know whether the thing causing your table experience is your own mind. Kvanvig (2003, 200), for example, offers an account of understanding according to which it is not only distinct from knowledge, but requires the “grasping of explanatory connections.” Of course, the veridicalist posits that we know the table’s causal role, including its connection to our experience. But how could we know how this causal role is connected to the table, if we don’t know what the table is? For illustration, consider a situation in which you know that there is a noise outside and you know how to make it stop affecting you – you can put in ear plugs. But you don’t know why the noise is happening or what its source is. In this case, it would be natural to say that, although you know what to do, you don’t understand what is going on, because you don’t understand why the noise is happening. Similarly, the veridicalist posits knowledge of how to manage your experiences and adjust your behavior, but without any understanding of why any of it is happening. To be sure, some might think that such a lack of understanding is simply a natural limitation of human cognition, or perhaps of the scientific enterprise. Constructive empiricists about science, for example, hold that a theory can be evaluated only for its empirical adequacy – how it matches observation – and nothing more (Van Fraaseen 1980). Accordingly, science cannot tell us why things are happening, in just the same sense that veridicalists cannot tell us why 26

Things are made more complicated here by the fact that, as Nick Treanor (2014) has pointed out, what looks grammatically like “one” fact can often contain a conjunction of many.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

26

Epistemology

things are happening. This suggests that the lack of understanding that veridicalism posits is no more troubling than the lack of understanding we are left with in science according to constructive empiricists. I take up comparisons between skepticism about whatsk things are and scientific humility in more detail in Section 4.2. For now, note that constructive empiricism is a rather skeptical view of what science can tell us about the world. (Van Fraassen himself was also skeptical about induction and inference to the best explanation.) And the general outlook that humans can never understand why things are happening around them is, of course, rightly regarded as a skeptical outlook. Let us return to the objection that knowledge of whatsk things are is not necessary for navigating around furniture, and set the notion of “understanding” aside for now. Consider other ways that sk might be significant for our lives besides navigating around furniture. Certainly, if I thought that I may well be in a vast simulation, this might affect lots of my thinking about my life, the difference I can make by my actions, and potentially death (whether there is more life once you “die” in the Matrix or in the demon’s mind is a different question from whether there is more life once you die on the more commonsense or “natural” hypothesis). Anyone who reflects on one’s life, and for whom these reflections matter in their life decisions and attitudes, will not be indifferent to sk, or whatsk things are. This is quite independent of whether we call such reflections or their outcomes “understanding.” Admittedly, this is not a matter of navigating around furniture, but why should navigation be the only benchmark for epistemic significance? The veridicalist antiskeptical strategy itself does not provide an answer, since it is not a sort of pragmatism about epistemic value, and since skepticism is, after all, a problem concerning epistemic value. I take up this issue again below, in Section 4. The second point against appealing to the score is that it seems, on reflection, not to matter. I’ve argued earlier that our information about the world, or the depth of our knowledge about it, is not significantly improved by replacing (2) with veridicalism. Rather, merely the form of our ignorance due to (1) changes. This is so despite the fact that, according to veridicalism, more of our “ordinary” beliefs about what ordinary things there are turn out to constitute knowledge. So, the only benefit, if there is one, of more of our ordinary beliefs turning out to be knowledge is the number, or score, of things we got “right.” However, what was disturbing about (4), originally, is the implication for our cognitive relation to the world, or the information we have about it, rather than the score, or number of beliefs that turn out to be knowledge. So, it is hard to see why the score itself would matter. The veridicalist rejection of (4) is a matter of bookkeeping of the little knowledge we have, rather than an addition to the little knowledge about the world we have according to (1).

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

27

In other words, “what” we know about the world is not the same as “how much” we know about the world. “How much” we know depends on how we’re counting, or how we formulate what we know. True beliefs, when their contents are more broadly construed or more easily satisfiable, are easier to come by, but that is not due to the improved informational situation, or even shifting epistemic standards (as is the case with contextualist solutions to skepticism). Rather, it is merely a reinterpretation of what our beliefs are about.27 To illustrate these points, consider an analogy. Suppose that you are concerned about your knowledge about the world due to your failure to know the exact color of any of the dinosaurs – there are educated guesses out there, and well-supported hypotheses, but, not being an expert, you don’t regard yourself as knowing. Then, your friend convinces you that, properly understood, birds count as dinosaurs. (Never mind whether this is correct, just suppose you are convinced.) So now, you think to yourself, you do know the color of some dinosaurs. Should you no longer be concerned? It depends on what concerned you about not knowing the color of any dinosaurs before. If it was that you failed to know the color of velociraptors, brontosauruses, and the other large reptilian things that you thought of as dinosaurs, then surely you shouldn’t feel any better now. You count yourself now as having more items of knowledge concerning the color of dinosaurs, but this doesn’t improve your knowledge about the things that concerned you before. The same is going on with veridicalist rejections of standard skepticism about what there is. For what concerned us, intellectually, about lacking knowledge of what there is – basic skepticism – is still the case, even if we can count more of our beliefs as items of knowledge about “tables.” If one was worried about one’s knowledge about the world on the basis of (1)–(4), one must have been assuming a false metaphysics of things like tables, according to veridicalism. Similarly, the worry about our ignorance of the color of dinosaurs was based on a false assumption about what counts as a dinosaur. But whether correcting these assumptions assuages the corresponding worry depends on whether the worry itself depended on the assumption, or whether instead merely the expression of that worry, as a worry about tables and dinosaurs, depended on those assumptions. And in both the skepticism and dinosaur cases, it is the latter. Expressing the skeptical worry as a matter of what there is rather than a matter of whatsk things are is a result of assuming a standard, nonveridicalist metaphysics of tables. But the root of the worry, basic skepticism, remains. Learning that birds are dinosaurs doesn’t assuage the 27

See Treanor (2014) for a more nuanced discussion of this general point about counting true beliefs.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

28

Epistemology

worry about brontosauruses, and learning that demonic ideas can be tables doesn’t assuage the worry that we don’t know whether the cause of our table experiences is an evil demon. Rather, learning veridicalism should lead whoever was worried about (1)–(4) to reformulate the worry as one about whatsk things are, or, more plainly, basic skepticism. And all you need in order to motivate that worry is (1). Let us sum up the case against the veridicalist antiskeptical strategy. If veridicalists are right about tables, then skepticism about what there is (and (4)) misdescribes our ignorance, or mis-expresses the worrying implication of basic skepticism.28 Given basic skepticism, which veridicalism accepts, the remaining question is not whether there is a table, but whether there is a certain kind of table here, a nondemonic, nonsimulated table. Why bother with such a question? Why should we reformulate the worry about what there is rather than abandon it? If one was ever worried about our knowledge about the world due to (1)–(4), then one should ask the new question because the veridicalist answer to the old question doesn’t assuage the worry, it merely changes its terms. Basic skepticism, the root of the original problem, is still there. If (4) was worrying because it suggested that we don’t know much about the world (as Stroud’s formulation suggests), swapping (4) for skepticism about whatsk there is doesn’t help, because it doesn’t improve our knowledge about the world. Rather, the terms in which the little knowledge we had (about the cause of our experiences) is expressed changed when we accepted veridicalism. We didn’t realize, before accepting veridicalism, that ideas could be tables, just as we didn’t realize birds are dinosaurs when we asked what color dinosaurs are. The mistake committed by (1)–(4) was not a mistaken evaluation of the state of our knowledge, but a mistaken interpretation of “table.”

4 Objections and Further Implications In the previous section, I argued that someone who knows that there are tables (as the veridicalist claims) does not necessarily know significantly more about the external world than someone who doesn’t know whether there are tables (as the standard skeptic claims). This is so, I argued, if the one who knows there are tables also doesn’t know whatsk tables are. One might think my conclusion surprising, perhaps even implausible. Doesn’t knowing what there is show that you know something more about the world, and doesn’t failing to know whatsk 28

Craig (1990, 213) offers a similar diagnosis of Davidson’s veridicalism: it goes wrong in its “over-concentration on a single sceptical formula, ‘It could be that most of our beliefs are false.’” Craig leaves it at that, without arguing that there is a formulation that is just as epistemically devastating, or that showing that most of our beliefs are true does not suffice to substantially improve our epistemic position.

The Skeptic and the Veridicalist

29

things are constitute a less radical shortcoming than standard skepticism? In other words, doesn’t veridicalism at least make some progress? And isn’t this talk about what things are a change in topic from the classic, traditional skeptical problem anyway? In this section, I address these objections and draw out some further implications of the failure of veridicalism to solve the skeptical problem. The objections can be understood as a defense of the view that the standard formulation of the skeptical argument, (1)–(4): (1) (2) (3) (4)

I do not know that ~sk. If I do not know that ~sk, then I do not know that there are tables. I do not know that there are tables. I do not know that there are any ordinary things.

is an adequate expression of the skeptical problem, since avoiding (4) is, according to these objections, sufficient to solve the problem (4) raises.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

4.1 Knowing That This Is Reality According to some versions of veridicalism, we know that the table, regardless of whatsk it is, is part of reality. As Leibniz put it, whatever there is here is “real enough.” This suggests that our knowledge about the world we experience according to veridicalism is better than our knowledge about the world according to (4). According to (4), the world we experience could be a deception or an illusion; according to some kinds of veridicalism, the world we experience is known to be reality. As implied in the first sentence of this subsection, only some versions of veridicalism hold that we know that this is reality, even if we don’t know ~sk. According to Chalmers, for example, we can know that this is reality despite not knowing whether we are in some global simulation (or other sk’s). This follows from his so-called simulation realism (2022, 106), the view that “If we’re in a simulation, the objects around us are real and not an illusion.” This is based on various conceptions of “reality” on which something is real if it exists, has causal powers, is mind-independent, is nonillusory, and is genuine (see 2022, ch. 6 for details). And, he argues, a global simulation will meet all of those criteria – though other skeptical hypotheses may not satisfy all of those. So, according to Chalmers, the table is real and you can know that it is real even if you don’t know that this is not all a simulation. And that, one might think, is a significant bit of knowledge about the external world that the skeptic who accepts (4) denies. Note here that, on this view, it is simply not possible for the thing playing the causal role associated with a table to fail to be real (unless you are in a local or nonglobal skeptical scenario, which we are setting aside).

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

30

Epistemology

According to Valberg, in contrast, if you were to wake up from a dream or emerge from a simulation, you would find yourself in a “wider” horizon, one in which someone was sleeping and dreaming, or envatted in a virtual reality machine. What makes the wider horizon, into which the dreamer emerges, “reality” is that there is no wider horizon than that. On Valberg’s view, “reality” means something like “the widest horizon: . . . Dream skepticism . . . asserts that I cannot justify my belief that nothing transcends [this personal horizon] . . . that I cannot justify my belief that there is no wider horizon.” (107) The situation in which the table is in a dream is a situation in which your horizon, that in which the table appears, is not reality (see Section 2). And this applies equally well to the BIV and demon hypotheses. So, whether or not this table is real depends on which scenario you are in, and since no one here is disputing (1) (I do not know that ~sk), we are assuming that you do not know which scenario you are in. So, you don’t know that the table is real. So, who is right about reality, Chalmers or Valberg? It is helpful to frame this question as one about the semantic neutrality of the notion of reality: either the property of reality (or being part of reality) is neutral (i.e. which property it is does not depend on whether you are in a skeptical scenario) or not. If it is, as Valberg suggests, then it is possible to be wrong about whether the table is real, even granting veridicalism. Only on a view like Chalmers’, on which real tables are simulations if you’re in a simulation, and they are ideas in a demon scenario, and so on, can the veridicalist claim to know that the table is real, or that what we experience is reality. One reason in favor of Valberg’s neutral account of reality is that, if this is all a simulation, then you might emerge from it, or be taken out the vat and detached from the computer running the program. And then you would be apt to judge that none of what you had experienced before was ever reality, even though you thought it was during the simulation. Intuitively, such judgments are true.29 The Chalmers-style veridicalist might reply that the meanings of words, such as “tree” and “real,” change to match the actual environment as soon as one emerges into the new environment, so that once you emerge, it is now true, but wasn’t just a moment before your emergence, that the world you were experiencing isn’t real. But in order for this rebuttal to succeed, it must imply that when you were in the vat, for example, your sentence “the things I experience 29

Chalmers does not apply his general veridicalist strategy to scenarios in which you have been recently envatted or put into the Matrix, but rather only “global” scenarios. So presumably he does not mean to apply his view to scenarios in which you were recently de-vatted, either. However, we can still consider how a subject would or should react if a subject were suddenly removed from its vat, or simulation. Our intuitions about the truth conditions of “what I experienced as a table was a real table” if that were to happen are consistent, of course, with the hypothesis that it will never happen.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

31

are real” expressed a true proposition, even though when you emerge from the vat, your sentence “what I experienced while in the vat was not reality” is also true. This would require the meaning of “real” to change suddenly, the instant you emerge. There is some question, which I cannot settle here, about whether our meanings can change so quickly due to our being in a new environment.30 But, furthermore, it seems that we are just as apt to judge, after emerging from the vat, that when we thought, before, that the things that apparently happened were reality (or “really” happened), those thoughts were false. So, on the current reply, we’d have to be mistaken about what we meant before, which is at least some significant cost to such a view of meaning. But perhaps the preceding difficulty for the Chalmers-style view is surmountable. Suppose instead that “reality” is not a neutral term, so that, like “wooden,” it applies to whatever realizes the right causal role. If so, then knowing that this is reality does not amount to much. For, “real” applies even to simulations (if you are in one). “Reality” is merely a label we can place on whatever explains why our experiences have a certain regularity. But knowing this is no better than what you can know consistently with (1)–(4), since even if (4) is true, you know that something explains why our experiences have a certain regularity (it could be a demon, robot overlords, etc.). So now it is hard to see why the ability to claim knowledge that what you experience is “real” is any achievement at all. This is not knowing much, since even a skeptic knows the same thing, even if she doesn’t use the term “real” to describe it. Moreover, this view seems to imply that, upon emerging from the vat, one could truly state, “I’ve been in reality all along, and still am now.” This seems intuitively false. Granted, on the Chalmers-style view, if you are currently in a simulation, and decide to create and run a simulation within your simulation, you may rightly regard the simulation you created as nonreal, while the world appearing around you – also a simulation – is reality on the current, nonneutral understanding of “reality.”31 So, the notion of reality can still do some work and make some relevant distinctions. But the point remains that, if “reality” is not neutral, so that it applies to your situation even if you are in a simulation, then knowing that this is reality is not a significant advantage over what you can know according to (1)–(4). For “reality” applies to whatever appears to be happening around you (so long as this scenario is global and stable), even if it doesn’t apply to a simulation you create within your environment. Either way, then, the claim that this is reality doesn’t show that the veridicalist can solve the problem raised by (1)–(4). 30

31

See Burge (1988) and Boghossian (1989) for some discussion of related “slow switching” cases, in connection with semantic externalism. This is so according to what Chalmers calls “realizer” structuralism.

32

Epistemology

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

4.2 Is Ignorance about Whatsk Things Are Merely Scientific Humility? It is undeniable that swapping out the standard metaphysics of things like tables does something to the skeptical problem: at the very least it changes the form of what it is we lack knowledge about, from what there is to what things are. And one might think that we have thereby reduced, or at least deescalated, the worry. Why regard ignorance of something about tables to be on a par with ignorance about whether there is even a table there in the first place? Moreover, the aspect of the table about which we remain ignorant according to verdicialism – whether it is ultimately an idea, or a computer simulation – is just something about its ultimate nature, and who among us ever seriously thought that we can know the ultimate nature of things? That is, one might suggest, merely a bit of appropriate and reasonable humility about the nature of things, a far cry from the skeptical disaster of ignorance about whether there are tables! Surely, then, the ignorance that veridicalism posits is altogether more acceptable than the ignorance that skepticism posits. Veridicalism implies mere humility, not radical skepticism. We already find scientific ignorance about the ultimate nature of things acceptable. That is, our current science is incomplete and is fully compatible with such ignorance. According to veridicalism, we know there are tables, but we don’t know what their ultimate natures are. And likewise, we once didn’t know the “H2O” hypothesis: that water is H2O. This was not a skeptically bad situation. Rather, one merely wondered: is water H2O, or is it fundamental as Thales thought, or is it something else? That is not skepticism, that is natural humility about the nature of things. And still today there is much we don’t know about what lies behind the manifest image. Consider what Chalmers calls “The Computational Hypothesis”: physics as we know it is not the fundamental level of reality . . . Underneath the level of quarks and electrons and photons is a further level: the level of bits. These bits are governed by a computational algorithm, which at a higher-level produces the processes that we think of as fundamental particles, forces, and so on . . . the Computational Hypothesis may lead us to revise a few metaphysical beliefs: that electrons and protons are fundamental, for example. But most of our ordinary beliefs are unaffected. (Chalmers 2005, 5–6)

Let us call the computational hypothesis “ch” for short. Chalmers (2005, 10) suggests that the Matrix (or BIV) scenario is an instance of ch, so ignorance of sk is no worse than ignorance of whether ch is true (as some physicists apparently hold).

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

33

What makes ignorance of whether sk any worse than ignorance of whether ch or H2O? They all concern what things are beyond how they appear, but ignorance of sk entails much more about our epistemic situation than ch and H2O. Unlike ignorance of ch or H2O, ignorance of sk implies that one can never, even in principle, know the fundamental nature of tables and water. The considerations that show that one currently does not know some scientific hypothesis like H2O concerns the current state of one’s empirical investigation, and this is compatible with one’s knowing some other things about the fundamental nature of things, at least at some other time and in other circumstances. In contrast, one’s failure to know whether sk arguably shows that one can never, even in principle, know the fundamental nature of tables and water, other than that they are whatever explains various features of our experience. For the empirical data required to know which hypothesis about the fundamental nature of the table is correct is presumably undermined by the failure to know whether, say, evil machines are generating all the empirical data in order to manipulate us. According to veridicalism, this table is ultimately computational if you’re in the Matrix. But the Matrix robot overlords could easily mislead you into thinking that the table is noncomputational by giving you experiences as of a Cartesian demon gradually “revealing” itself to you as the source of your experiences. And, presumably, the robot overlords could also mislead the physicists in the Matrix themselves, whatever theory the physicists accept on the basis of empirical data. Whatever scientific evidence you claim to have about the ultimate nature of things, presumably robot overlords, BIV machines, or demons could have planted that evidence to mislead you. The same holds, of course, for H2O and the precise chemical composition of water. If you’re ignorant of sk, then science, when it comes to the ultimate nature of things, is not just incomplete, it is hopeless. It follows that ignorance about sk is not just as acceptable as ignorance about ch or H2O. For, what makes ignorance about ch and H2O acceptable is that investigation is ongoing, that the scientific enterprise at any given time (at least before it becomes “complete”) can’t be expected to tell us everything about the fundamental nature of things. It is reasonable to accept the relative immaturity of the science of fundamental nature. But if we are ignorant of sk, then the immaturity of the scientific enterprise is irrelevant, since we can never know what things ultimately are. Science can’t be expected to tell us anything about the fundamental nature of things. This is worse than the truism that science can’t tell us everything. To be clear, some scientific hypotheses can arguably be known even if we are ignorant of sk, given verdicailsm: those expressible in nonneutral terms. We can know that the table is made of wood, and that wood comes from felled trees,

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

34

Epistemology

and that trees photosynthesize, given that these are nonneutral terms. So scientists can investigate such claims regardless of our ignorance of sk. Of course, if this were all happening in the Matrix, machines might be able to trick the scientists into getting even those hypotheses wrong, since they can produce any sequence of experiences, or data, they want to. But this is a matter of messing with your life within the simulation in ways that are probably not the global scenarios veridicalists like Chalmers have in mind (such scenarios are more like The Truman Show than The Matrix). However, the fact that some science may be possible even if we are ignorant about sk does not address the main problem with the objection. The point is that trees and photosynthesis may be, according to (1), ultimately demonic ideas. This sort of ignorance is of a different sort, as I have suggested, than what mere scientific humility suggests. Hypotheses about the fundamental nature of things that are expressed in neutral terms are unknowable, if sk is unknowable. (On the Putnamian view considered earlier, on which there are no neutral terms, there are no such hypotheses; but that is not veridicalism, and it amounts to a rejection of (1).) Nor does the preceding imply that there is some clear-cut principle distinguishing scientific from skeptical hypotheses. One might regard it as acceptable scientific humility that we can never know, say, whether this universe is embedded within a larger one, or even whether there is a level of reality “under” the ones that we can discover, which is entirely different from what we could ever know about the inner nature of things. And if we cannot know whether the universe is infinitely complex, perhaps we cannot know whether there is any fundamental level of explanation at all. Such humility seems reasonable, and, like skeptical ignorance of sk, is an in-principle ignorance rather than a contingent one based on the present state of the scientific enterprise. Insofar as such hypotheses undermine some of our basic assumptions about what things are, or what the world is, they may well qualify as instances of sk. (This seems not to be the case with the hypothesis that our universe is embedded in a larger one, nor with the hypothesis that the universe is infinitely complex, since those do not seem to imply anything about the reality or not of the things we see in this universe, at least not like the simulation hypothesis intuitively does.) Though, insofar as these are beyond our ability to confirm or disprove empirically, it is also unclear to what extent these are parts of scientific inquiry at all. Perhaps it will remain unclear, then, whether some hypotheses, even if they count as “scientific” because they involve some scientific concepts, are possible instances of sk. Perhaps ch is one such unclear example. This doesn’t show that our ignorance of sk is mere scientific humility; rather, it shows that some

The Skeptic and the Veridicalist

35

scientific hypotheses are so strange that, given that they cannot be ruled out, they can serve as skeptical hypotheses. Science fiction has, of course, supplied much of the material for modern skeptical hypotheses, so this is hardly surprising.32

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

4.3 Is Ignorance about Whatsk Things Are Merely Metaphysical Humility? One might think that it is not the job of science to tell us the ultimate nature of things, so that ignorance of ch (construed as an ultimate nature hypothesis) is not scientific but metaphysical ignorance. But then, even so it is perfectly acceptable and reasonable to be humble about such ignorance. That is, we are also (arguably) ignorant about the world due to the incompleteness of metaphysics, and this is no skeptical disaster. So why is skepticism about whatsk things are any different? Isn’t that just more metaphysical humility? I have two replies: metaphysical views are quite different from skepticism, even while our ignorance of which metaphysical view holds may in fact be quite similar to skeptical ignorance. Second, that latter point helps emphasize that veridicalism can help us to better understand the nature of skepticism as involving metaphysical issues, not just epistemological ones. If you are a demon victim, objects are mind-dependent. But some metaphysicians endorse panpsychism, as Chalmers (2005, 18–19; 2018, 20) points out. Panpsychism is not considered a skeptical scenario, yet it also entails that objects are mind-dependent, at least in the sense that if not for some mental state (i.e. their own), they would not exist. Likewise, according to veridicalism, the BIV scenario has it that tables are ultimately computational. But so does ch. Suppose we are ignorant of whether such hypotheses as panpsychism and ch are true (and suppose we treat ch as a metaphysical rather than a scientific hypothesis). Why is this not disturbing in the way that ignorance about sk is? Here, one cannot appeal to the fact that skeptical scenarios are (purported to be) unfalsifiable. For it could be argued that metaphysical views such as panpsychism are also unfalsifiable. If so, then why couldn’t panpsychism be a version of sk? Like Berkeley’s idealism, panpsychism is a positive view about what things ultimately are. Skepticism about whatsk things are is a negative view about our ability to know any such theory, or any of its alternatives. It may well be that our inability to know whether panpsychism is true demonstrates a significant inability of ours, namely to know the nature of things. If so, then panpsychism could, being employed in this way, serve as one version of sk. Whether panpsychism is as effective at highlighting our ignorance as the Matrix is depends on how deep 32

For related discussion on quantum phenomena in relation to skepticism, see Vogel (2019, 9–10).

36

Epistemology

we judge the commonsense assumption against panpsychism to be. And this may be somewhere on a sliding scale. There may not be a clear-cut answer to the question, whether ignorance of whether the table is a demonic idea is as disturbing, in its epistemic consequences, as ignorance of whether the table is conscious (as according to panpsychism). They are both items of ignorance about what the table is (according to (1)), and both clash with our assumptions about the world to some degree. Since it seems, to my estimation, that the demon scenario clashes with our commonsense assumptions more spectacularly, it seems that the demon is a better scenario to use for the purpose of demonstrating our inability to know about the world. Intuitively, it seems worse to be ignorant of whether you are a demon victim than it does to be ignorant of whether all of nature is mind-independent, because the former seems farther out relative to our usual assumptions. A similar point can be made about “Ramseyan Humility” (Lewis, 2009), which Lewis describes as

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

ignorance of the intrinsic properties of substances. The substances that bear these intrinsic properties are the very same unhidden substances that do indeed affect us perceptually. But they affect us, and they affect other things that in turn affect us, in virtue of their causal powers, which are among their relational properties. Thereby we find out about these substances as bearers of causal powers, but we find out nothing about them as they are in themselves. (2009, 203)

In other words, we can never know what substances in themselves, which have the structural properties that we interact with, are. They could, for all we know, be a BIV machine, a demon’s mind, and so on. But they could also be parts of God’s mind, or a quantum wave function, or some other, non–classically skeptical fundamental substance. Lewis did not consider Ramseyan humility to be a sort of radical skepticism, or an epistemic disaster as implied by (1)–(4). He wrote, “who ever promised me that I was capable in principle of knowing everything?” (211). But he did not explicitly have in mind things like the demon or BIV scenarios. And, again, it seems that ignorance of the BIV scenario is more disturbing than the other possibilities, because it clashes with our basic assumptions about life more dramatically. For example, you probably assume that your “brain,” here in this experienced space, determines your experiences. Ignorance of whether you are a BIV undermines this, since the thing floating in a transcendent vat, having these very experiences stimulated by something that is not even spatially related to the table that you are experiencing, might have equal claim to being the thing that gives rise to your experiences – it certainly would be a background condition that must be in place for your virtual brain to be “causing” your experiences. Some

The Skeptic and the Veridicalist

37

possible substances that bear causal powers, or “realizers,” clash with our assumptions about the world so violently that our ignorance about them disturbs us and undermines what is intuitively our knowledge about the world. As Rae Langton put it,

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Something or other realizes the role of provider of sensory experience. This something could be a material world; or it could be a Demon and his machinations . . . .Unless we can properly ignore some alternative possible realizers of this role, knowledge of the real world will disappear. (2004, 134, emphasis added)33

There are some ways of filling in the details of what substances in themselves are that make too much of a mockery of our assumptions about the world. “Who ever promised me that I was capable in principle of knowing whether this table is an idea in someone’s (maybe my) imagination” does not have the same comforting ring to it as Lewis’s original. This is because that is not an expression of humility, it is an expression of skepticism: you have no idea what is going on around you. It should be conceded, though, that there is no bright line between skeptical scenarios and such metaphysical hypotheses. This corroborates the point, made earlier, that veridicalism helps to establish the metaphysical subject matter of the skeptical problem: we can never know what things are. Still, this does not seriously undermine the main concerns about veridicalism as a response to skepticism. For, just as not every sort of ignorance about what things are constitutes skepticism, nor does every sort of ignorance about what there is constitute skepticism. Suppose you wake up in a basement one day, after a near-apocalyptic war, wondering if any furniture in the world has been left undestroyed: “Are there any tables?” This would not be a case in which you are suddenly in a skeptically bad situation of ignorance about what there is. And yet, even veridicalists agree that (3) is skeptically disturbing. The fact that some other ways of being ignorant of what things are fail to be skeptically disturbing doesn’t show that all ways of being ignorant about what things are fail to be skeptically disturbing.

4.4 The External World versus Other Minds Here is another way to argue that ignorance of whatsk things are is not as bad as the skeptical worries brought on by (4).34 Suppose we come up with a view on which there are other minds, but which does not settle whether minds are 33

34

Granted, Langton ultimately rejects this skepticism, but the passage quoted here indicates that she took ignorance of what ultimately generates our experiences as a potential source of skeptical worries about the external world. This was suggested to me by an anonymous referee.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

38

Epistemology

physical, so that we remain ignorant of “what minds are.” Surely this would solve the “problem of other minds,” even if it leaves open the nature of minds. Why isn’t the view, offered by veridicalism, that there are tables, but which leaves open whatsk tables are, just as satisfying a solution to skepticism? The answer is that veridicalism leaves open much more about the nature of tables than this view about minds does in leaving open whether minds are physical. What’s worrying about the problem of other minds is, primarily, that we don’t know whether other people are conscious. It is a certain conception of other minds – as conscious – that enables us to express the worry as one about minds. Likewise, what is worrying about the existence of tables, at least as it comes up in the context of skepticism, is primarily that we don’t know whether all things in the external world (that is, beyond our sensations) are ideas, or simulations. It is a certain conception of tables – as nonsimulations and nonideas – that enables us to express the worry as one about the existence of tables. The view that minds exist but may not be physical still addresses the question of consciousness, and so it solves the problem of other (conscious) minds, so long as we accept that something physical could be conscious, of course. The view that tables exist but might be ideas does not address the question of whether tables are ideas, so it does not solve the skeptical problem. Rather, it directly challenges the conception of tables that was assumed when the problem was expressed as one about the existence of tables. So, a better analogy to veridicalism about tables is the view that there are other minds, which leaves open extreme eliminativism about consciousness. That is, on this alternative view, there are minds but they might not be conscious. There is no problem of other “minds,” but only because this radical view about minds makes it much easier to count as a mind than it is according to the standard view that is assumed by those who usually pose the problem as one about “minds.” On this alternative view of minds, even zombies and robots can count as having minds, so the problem of other “minds” is better expressed as the problem of other “consciousness.” This is what bothered us about other minds in the first place, of course, and is a reformulation of the original problem. For this reason, the alternative view of minds that leaves eliminativism open is not itself a solution, but a rejection of the terms of the problem. Not just any theory of mind is such that, as long as we end up knowing that there are other minds, the problem of other minds is thereby solved. Such a theory must be consistent with the conception of minds that led us to formulate the problem as one about “minds” in the first place. What is the criterion for what must be assumed about minds in order for such a theory to constitute a solution? Minds need to be conscious. Otherwise, what we were originally worried about in the problem of other minds will not have been addressed. We need to consult the

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

39

assumptions about what minds are that made us express the worry as one about minds in the first place. Likewise, our worry in (3) about the existence of tables depended on the assumption that tables cannot be ideas, or simulations. Once that assumption is removed, the claim that there are tables no longer assuages the original worry, but rather it changes the terms. So, the view that there are tables, but they might be ideas or simulations, doesn’t help, given that what one was worried about in (3) was that it implied that the things we call “tables” might be mere ideas. One might reply: what about a view according to which other people have consciousness, but that doesn’t settle whether consciousness is physical? That would indeed solve the problem of other minds, assuming we can know such a view to be true. But then, similarly, a view according to which there are tables, which doesn’t settle the fundamental physics of such things as tables, solves the skeptical problem, too. It’s just that this is not equivalent to a view, like veridicalism, that determines that there are tables but not whether they are BIV simulations, as I argued earlier. When we worried about other minds due to the inaccessibility of others’ experiences, we assumed minds have consciousness, not that they are physical (or nonphysical). When we worried about tables due to (1), we assumed tables aren’t simulations, not that this or that other hypothesis about the nature of matter applied. Show me that other people are conscious, and I’m satisfied about other minds. Show me that tables, or whatever causes my table experiences, are not simulations, and I’m no longer worried about (4). But veridicalism doesn’t do that. Only a refutation of (1) (not just (2)) can do that.

4.5 It Doesn’t Matter Whatsk Things Are If things look and behave just as if there are nonsimulated tables, then why does it matter whether we know that they are nonsimulated tables, especially if, as veridicalism states, we know that they are tables? I’ve been arguing that it matters because it implies that we don’t know much about the world, and specifically we know no more about the world than we do according to (4). The objection here is that it doesn’t matter in some other way. If so, then at least veridicalism significantly lowers the stakes of skepticism. It will help to first address an idea that might muddy these waters. One might claim, as in Chalmers (2005, 16), that if you learned that you are a BIV, or deceived by a demon, then you’d get used to it and life would go on. If so, it might be thought, then who cares whether those scenarios are happening?

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

40

Epistemology

And given that it wouldn’t matter to you much (pragmatically, or otherwise), it must not matter whether sk is true. So, the thought goes, ignorance of whether sk is true and whatsk things are is just not such a significant kind of ignorance, certainly not as significant as ignorance of what there is. However, there is a difference between learning that some strange, unexpected hypothesis like sk is true, versus not knowing which of a wide range of hypotheses is true. Only the latter is the sort of ignorance characteristic of skepticism. The skeptical problem is not the problem of knowing that some unexpected or wild scenario is going on. Rather, it is the problem of ignorance about which scenario is going on. That you would get used to learning that, say, the evil demon hypothesis is true does not show that lacking knowledge about whether that is so is insignificant. Nor does it show that it doesn’t or shouldn’t matter very much which possibility holds. Consider some analogies: I would get used to it if my best friend died, and life would go on. That doesn’t imply that it doesn’t or shouldn’t matter whether she dies. And it doesn’t make ignorance of whether she has died any less significant. Likewise, suppose you don’t know whether there is an afterlife, but you realize that if you learned that there is no afterlife, you may well get used to it (people often do), and life goes on (and then it ends). This doesn’t suggest that it doesn’t matter whether death is the end; it is still a profoundly important matter. And it doesn’t show that your ignorance of whether there is an afterlife is any less significant. The relevant question raised by skepticism is not how would you understand your place in the world if you learned that sk is true, but rather, how do you understand your place in the world, given that you do not know whether sk is true? With this point in mind, let us consider whether not knowing which scenario holds matters pragmatically or otherwise. 4.5.1 Pragmatically This version of the objection is that, so long as (3) is false, and I know I can put a cup on the table, sand it down if I need to, move it closer to the window, and so on, I can plan my day-to-day life, rationally, around this knowledge. So, it doesn’t matter that I am ignorant of whether the table is a dream table, or a simulation, and so on. However, it makes sense to care about things that affect oneself and others only insofar as the ways in which you try to affect changes are really the ways that those changes are affected. But none of this is so in a BIV scenario. For all you know, nothing that you try to do will make anyone else happy, as you want them to be – perhaps their vat machinery is programmed to ensure that they are always unhappy. In the Matrix scenario, nothing you do will alter the large scale

The Skeptic and the Veridicalist

41

course of history.35 This makes a pragmatic difference if your goals take anything like changing others’ lives into account. Thus, ignorance of whether you are a BIV can affect your plans for what to do and your attitudes about others, given your practical goals that involve them. Of course, everyone (even the veridicalist) agrees that some metaphysical hypotheses can have pragmatic significance, and traditional theism is a case in point. What I have been arguing here is that our ignorance about skeptical hypotheses, which are, according to veridicalism, also metaphysical hypotheses, can have pragmatic significance as well. It can make a pragmatic difference if you are ignorant of whether your life is a vast simulation.36

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

4.5.2 Other Values One might suspect that whatsk things are does not matter, but not because it makes no pragmatic difference. Rather, whatsk things are does not matter because there are some other nonpragmatic (and still nonepistemic) values that are indifferent to whether we know whatsk things are, or whether we know not-sk. Valberg (2007, 123–125), who argues that not knowing whether this is all a dream is indeed ultimately a significant skeptical problem, argues that (1) undermines all value. If this is all a dream, then everything that matters is internal to the dream. If so, then everything that seems to matter ultimately does not matter, and that seems to matter. If this is a dream, then you do have a family and a home. But, as we saw in Section 4.1, on his view these are not “real” family and home if this is a dream, any more than the family and home you may have once dreamt about are real. You think they matter only when you think they are “real.” If you do not know whether this is all a dream, then you do not know whether everything that seems to matter does not matter after all.37 The same thought applies to BIVand demon hypotheses – it is important to note this, since some might exclude dream hypotheses from sk, for reasons given earlier. Perhaps, as I noted earlier, if you somehow learned that you were in the Matrix, you would get used to it and start caring about the virtual things in your environment. But getting used to such conclusions is one thing; whether things inside the Matrix really matter is another. Given your state of ignorance according to (1), you might be mistaken or ignorant about the nature of what matters. You might think your inherited table matters, but its nature, or whatsk it 35 36

37

See Pryor (2005). See Schwitgebel (2017) for more examples of the practical consequences of being uncertain about skeptical scenarios. But see Chalmers (2017) for some discussion of the value of virtual objects, and Chalmers (2022, part 6) for general discussion of values in a global simulation.

42

Epistemology

is, might be entirely different from what you take it to be, and so might not matter after all.38 If you find nothing in the world, apart from your experiences, to be valuable at all, then of course your ignorance of whether sk is true will not make a difference to what you value. The present point is that if you value something in the world external to your experiences, then ignorance of whatsk things are, and of whether sk is true, will matter to that thing’s value. And this is because whatsk things are is a matter of their nature external to how they affect your experience.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

4.6 A Historical Worry: Have I Changed the Topic? So far I have argued that the sort of ignorance due to (1) that veridicalism posits – of whatsk things are – matters about as much as standard skepticism due to (1), and that it posits the same ignorance about the external world. So, rejecting (4) while accepting veridicalism and (1) does not solve the skeptical problem posed by (4). But one might think that, even so, skepticism about whatsk things are is a new problem, or a change of topic, so that veridicalism still puts to rest skepticism as it has always been conceived. The issue here is whether the skeptical problem ought to be identified with (4). Since this issue concerns a long philosophical tradition, it calls for some historical discussion. Since I cannot discuss the entire vast tradition here, this discussion will be limited to supporting this claim: one can find historical formulations of skepticism about the external world that can be reasonably interpreted as being compatible with not-(4) and veridicalism, since they invoke skepticism about whatsk things are, rather than exclusively skepticism about what there is. To be clear, I do not claim that any of these thinkers were veridicalists, or even had veridicalism in mind. Rather, they arguably formulated skepticism in such a way, unwittingly or not, as to make veridicalism a moot point. Ignorance of whatsk things are – their inner, intrinsic, or true nature – is arguably one sort of ignorance the ancient skeptics, described by Sextus, had in mind. The ancient skeptics accepted “appearances,” but withheld judgment on anything about the appearances, or about what lies behind them, or their true nature. The claim that we don’t know what lies behind appearances is compatible with there being something here that looks and behaves like a table. Tables, according to veridicalism, are things that appear a certain way to us and behave a certain way, regardless of what their underlying nature is. It is, arguably, ignorance of that underlying nature that the ancients were concerned with; whether they thought that this underlying nature determined that this table 38

See Button (2013, ch. 16.6) for some relevant discussion of values and skeptical scenarios.

The Skeptic and the Veridicalist

43

appearance counts as a table seems to be left open. At least part of the ancient skeptical worry, then, concerned the nature of the things behind appearances that we remain ignorant of due to (1), even granting veridicalism.39 The representative of skepticism in early modern times, as it is known to most contemporary analytic philosophers, is Hume. In the Enquiry, section 12, part 1, he takes the skeptical problem regarding our senses to be the following:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

By what argument can it be proved, that the perceptions of the mind must be caused by external objects, entirely different from them, though resembling them (if that be possible) and could not arise either from the energy of the mind itself, or from the suggestion of some invisible and unknown spirit, or from some other cause still more unknown to us? It is acknowledged, that, in fact, many of these perceptions arise not from anything external, as in dreams, madness, and other diseases. And nothing can be more inexplicable than the manner, in which body should so operate upon mind as ever to convey an image of itself to a substance, supposed of so different, and even contrary a nature . . . .It is a question of fact, whether the perceptions of the senses be produced by external objects, resembling them: how shall this question be determined?

The question is about what causes our perceptions, and the challenge is to show that the cause is not “energy of the mind itself,” “some invisible and unknown spirit,” or some other cause unknown to us. In other words, the challenge is to show that our perceptions are not caused in some sk-type way. Whether this amounts to showing that there exists something that we can call a table depends on whether Hume was assuming that, on such skeptical possibilities, there are no tables. But Hume doesn’t make any such assumption explicit in his description of the skeptical worry. It follows directly from (1) that whatever causes your table experience might be your own mind, or a demon, or some other unknown source, and that is what Hume explicitly identifies as the problem of skepticism. Veridicalism is fully consistent with this. This is what I earlier called basic skepticism, which follows from (1) directly, even according to veridicalism.40

39

40

See Burnyeat (1982, 25–26). Of course, the ancients were ignorant of some scientific facts about the nature of things, which veridicalism allows that we can know. So, the claim here is not that there is no difference between what the skeptics claimed we are ignorant of and what veridicalists allow that we are ignorant of. Rather, the claim is that part of the ancient skeptical claim was about the nature of things that we are ignorant of due to (1), and we are ignorant of this even granting veridicalism. After the passage quoted here, Hume continues: “Bereave matter of all its intelligible qualities, both primary and secondary, you in a manner annihilate it, and leave only a certain unknown, inexplicable something, as the cause of our perceptions; a notion so imperfect, that no sceptic will think it worth while to contend against it.” This further suggests that Hume thinks that the beliefs vindicated by veridicalism do not include the beliefs his skeptic threatens.

44

Epistemology

Besides Leibniz, quoted at the outset of this Element, Berkeley might be thought to be the closest thing to a veridicalist in early modern history. According to his idealism, tables exist but are ideas in the mind of God. Famously, he insisted that he is not a skeptic but rather a champion of common sense, which may suggest he is an antiskeptical veridicalist. If so, then perhaps he also equated skepticism with (4). But that would be a misunderstanding. Berkeley defined “skepticism” as the view that we are ignorant of the inner nature of things. In the preface to the Dialogues, and speaking as himself rather than a character, he writes:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

“Hence arise skepticism and paradoxes. It is not enough that we see and feel, that we taste and smell a thing. Its true nature, its absolute external entity, is still concealed” (Works, vol. 2 167).

Since he thought that we (or at least he and his followers) are not ignorant of the inner nature of things, he did not consider himself a skeptic. We know what tables are: they are ideas. But if, as veridicalism plus (1) has it, we cannot know whether tables and other such things are ideas, skepticism triumphs after all. As Pappas (2008, 251) puts it in a discussion of Berkeley, “One may have all manner of knowledge of objects, including that of their existence, but remain ignorant of their ‘inner’ nature.” This is exactly the state Berkeley thinks of as skeptical, though he did not consider himself a skeptic. The point is that his formulation of skepticism was compatible with veridicalism.41 So, veridicalism itself, since it is consistent with (1) and our ignorance of whatsk things are, would not refute some historically important formulations of skepticism. The appeal to the skeptical worry raised by those formulations, that we are ignorant of whatsk things are, is not a change of topic, but rather a reminder that the skeptical problem raised by (4) can be and has been raised in other ways. Among the mainstream historical sources of contemporary skepticism, G. E. Moore, in the twentieth century, who focused as much on idealism as on skepticism, seems to be the first to explicitly and exclusively focus on what ordinary things, such as tables and hands, there are, or (4). I speculate that it is no coincidence that Moore himself was open to a sort of veridicalism. In In Defense of Common Sense, he distinguishes truisms that are understood by any competent speaker (e.g. “there are physical objects”) from the correct analysis of those truisms (e.g. that physical objects are made out of sense-data). Since the truisms are common sense and accepted even by skeptics and idealists (even if unwittingly), there is no question about their truth. But what they are about, 41

In the dialogues, Hylas defines a skeptic as “one that doubts everything.” But he soon after proposes a reformulation: “distrusting the senses . . . denying the real existence of sensible things, or pretending to know nothing of them” (Berkeley and Mathias 2007, 173).

The Skeptic and the Veridicalist

45

and their logical form, is up for debate. This gives our ordinary beliefs the sort of immunity from error that Chalmers, for example, defends, but leaves open what the correct metaphysics is, or what their intrinsic properties are. Perhaps he equated skepticism with whether there are ordinary things such as tables because that was the problem he thought he had a solution to.42 But before Moore, it is hard to find an explicit identification of skepticism with (4). Even Descartes, whose name we borrow when we speak today of “Cartesian skepticism,” despite the fact that he was not a skeptic, did not always distinguish skepticism about what there is from skepticism about whatsk things are. That is, he did not clearly distinguish, and so presumably did not think it made a crucial difference, between whether there are tables and whether tables are mere ideas. In the final paragraph of the first meditation, which introduces most students to the topic of skepticism, and which is the most often cited source for Cartesian skepticism, he wrote:

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

“I shall consider that the heavens, the earth, colours, figures, sound, and all other external things are nought but the illusions and dreams of which this genius has availed himself in order to lay traps for my credulity . . .”

Does this mean that there are no external things, or that external things are illusory, or dream-objects? This sentence does not require us to interpret either way. Only after that passage does he interpret this to mean that there are no such objects.43 Clearly, the difference between external things being illusions, and there being no external things but instead only illusions, was not essential to Descartes’ thinking about skepticism; they seemed, judging by his writing, to amount to the same problem. But veridicalism would only address one of those, namely whether there are tables, and not whether those tables are mere illusions. Skepticism about whatsk things are is not a new issue in the history of skepticism or a change of topic. It has appeared, alongside skepticism about what there is, as part of the focus of skeptical worries throughout the skeptical tradition. Only basic skepticism, about what it is that causes our experiences, seems to be dominant throughout the history of skepticism about the external world.

5 General Implications for Skepticism and Antiskeptical Strategies At the beginning of this Element (in Section 1), I suggested that one can know a lot about which things exist, while knowing very little about the external 42 43

See Walker (2020) for discussion of Moore’s brand of veridicalism. Here I am quoting Elizabeth S. Haldane’s translation from The Philosophical Works of Descartes, 1911 (Cambridge University Press). “Nought but” is translated as “merely” in Cottingham’s more recent translation.

46

Epistemology

world. And I suggested that this has important implications for how we understand skepticism, what is at stake, and how we can best formulate the skeptical argument. Most of the preceding is pretty specific, in that it specifies what is wrong with a particular antiskeptical strategy – veridicalism – and considers various attempts to rescue that strategy. In the process, we have seen in detail that skepticism about whatsk things are can be just as worrying as standard skepticism about what there is, or (4): (4) I do not know that there are any ordinary things. As a result, veridicalism, which rejects (2): (2) If I do not know that ~sk, then I do not know that there are tables. and thereby avoids (4), fails to solve the problem posed by (4), since it commits to skepticism about whatsk things are, which is as problematic as (4). In this final section I explain how, from that specific discussion of veridicalism and skepticism about whatsk things are, we reach the more general and sweeping claims I made at the outset.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

5.1 Beyond Veridicalism to Other Rejections of (2) To begin, we can generalize the negative conclusion about the inadequacy of veridicalism to solve the skeptical problem. There are other ways of rejecting (2) while accepting (1) (I do not know that ~sk) besides veridicalism, such as the strategy of rejecting closure. On this strategy, knowing that there are tables does not require knowing, or even being in a position to know, that one is not a BIV. This is due to the nature of knowledge rather than the nature of tables.44 If we leave it at that, then it seems this strategy raises the same problems that veridicalism does. For if you do not know whether you are BIV, then even if you know that there is a table, you still do not know whether the table is part of a simulation. You don’t know whatsk it is. Any bare rejection of (2) that leaves (1) untouched implies skepticism about whatsk things are. This is not to say, however, that all rejections of closure will be committed to skepticism about whatsk things are. Some rejections of (2) come with commitments regarding (1), which, while they are compatible with (1), posit something further about sk. Some Wittgenstein-inspired views, for example Avnur (2012) and Coliva (2015), reject (2) but propose that in order to know things like that 44

Denying closure does not address dreaming hypotheses, which are clearly compatible with the existence of tables. However, insofar as (2) is plausible for such hypotheses, there must be some other plausible principle besides closure that specifies the relevant sort of relation between our beliefs in ordinary things and dreaming. Stroud (1984) is a classic discussion of the principles that connect knowing that p and knowing that you are not dreaming that p.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

The Skeptic and the Veridicalist

47

there is a table, one must assume (or have high confidence) that not-sk. This assumption does not constitute knowledge, and lacks evidential justification, but it must be in place in order for one to have knowledge of ordinary things. If so, then unless one assumes that one is not, say, a BIV, one cannot know that there is a table.45 Any situation in which one knows that there is a table, then, is also one in which one at least assumes that it is not a BIV-table. In other words, no one is in a position to correctly claim to know that there are tables, but leave open whether sk is true, so that the tables might be, say, BIV-tables. So, skepticism about whatsk things are does not arise as it does with veridicalism. However, not all denials of closure can avoid skepticism about whatsk things are so easily. For example, Dretske (1970) and Nozick (1981) seem to hold that no assumption, confidence, or belief (let alone knowledge) that not-sk is required for one to know that there is a table. A person is often in a position to correctly claim to know that there are tables while being entirely open to the possibility that sk. Such a person could wonder, then, whether the table is a BIVtable. For example, Dretske (2005, 16) discusses what he calls “heavyweight” implications of things we know: things that cannot be seen to be true, but which are entailed by things that can be seen to be true. On his view, we can fail to know such heavyweight implications, so that, for example, you can know that there is wine in the bottle while not knowing that the liquid in the bottle is not colored water. Applying the idea to sk, you can know that there is a table without knowing that it is not a BIV- or demon-table. So, his view implies skepticism about whatsk things are. So, on the question of whether rejecting (2) by rejecting closure raises the problem of whatsk things are, the verdict is mixed. Generally, any view that rejects (2) while leaving it open, from the subject’s perspective, whether sk is true, will have this problem.46 45

46

This is an oversimplification. As Coliva (2015, 35) notes, children or others who do not have the requisite concepts to assume anything about brains and vats still can know that there are tables. This is possible because such people are participants in a social-linguistic practice of evaluating beliefs which presupposes (or has as a “rational precondition”) the assumption that not-sk, so that anyone participating is, in virtue of participating, implicitly assuming not-sk. Pritchard (2016) may also fall into this category, though this is complicated by the fact that he claims that instances of sk and their negations – which he calls “hinge commitments” – are not objects of belief. This raises the question whether claims about whatsk things are (such as that the table is not a BIV-table) are objects of belief. The “uber-hinge,” which along with the other hinges is excluded from closure principles, is that “one is not radically and fundamentally mistaken in one’s beliefs” (95). Does this uber-hinge include (or entail) that tables are not BIVtables? If it does, then no one is ever in a position to correctly state that they know that there is a table, but not whether it is a BIV-table, so Prtichard is not yet committed to skepticism about whatsk things are. In footnote 5 (appearing on p. 203) to his definition of the uber-hinge, we get a clue as to whether this is the case. There, Pritchard clarifies that he does not mean to measure how fundamentally one is mistaken by the sheer number of true beliefs. So it seems that among the things that are not objects of belief are claims about whatsk things are, and skepticism about

48

Epistemology

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

5.2 All Successful Strategies Must Deal with (1) What this shows is that antiskeptical strategies must address (1) rather than focus exclusively on the relation between (1) and our ordinary beliefs about what there is. It appears that the exclusive focus on beliefs such as “there is a table,” considered “ordinary” even though they seldom come up explicitly in daily life, began in the twentieth century with G. E. Moore (see Section 4.5). But this is but one peripheral formulation of the worry that we know next to nothing about the external world, which is better expressed by what I earlier called “basic” skepticism: you do not know what causes your experiences. However, there is a remaining question, not addressed in the preceding sections, of whether skeptical scenarios such as those featured in (1) are themselves necessary for bringing up the skeptical problem. One could raise the basic skeptical worry without appealing to sk, or scenarios in which it is some nightmarish, alternative cause that gives rise to your experiences. For example, one might, as Hume once did, question what source of information one could have concerning the ultimate sources of one’s experiences (where, presumably, one’s experiences cannot themselves reveal their own source). Still, it seems fair to say that in order to solve the problem of skepticism, one must be able to deal with (1). If you can figure out, somehow, what the ultimate cause of your experiences is, then presumably you will be in a position to rule out that the ultimate cause is, say, a demon or a dreamer. Thus, though I reject the idea that skeptical scenarios are necessary for articulating the problem of basic skepticism – which is ultimately the root of all skepticism about the external world – if you can refute basic skepticism, you should be in a position to refute (1).

5.3 The Metaphysical Question and the Standard Formulation From the preceding discussion, it follows that the currently standard formulation of the skeptical argument is flawed. If one can refute the argument by rejecting (2), thereby avoiding the conclusion (4), and yet still be saddled by the problem that (4) gives rise to, then (1)–(4) is not an ideal formulation of the skeptical problem. I will not here offer an alternative, but rather diagnose the problem. We have seen that the reason rejecting (2) is insufficient to solve the problem raised by (4) is that one might be left with a skepticism about whatsk things are, which is just as problematic as (4). The only way to evade this problem is to endorse a standard metaphysics of objects like tables. That is, one has to assume whatsk things are is avoided. I leave it for another occasion to untangle the implications, if any, for the rest of Pritchard’s epistemology.

The Skeptic and the Veridicalist

49

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

that there are no tables in a global skeptical scenario in order to avoid this problem with the standard formulation. It follows that, insofar as the standard metaphysics is questionable – and the veridicalists have done an impressive job showing that it is – so is the standard formulation of the skeptical argument. While this doesn’t solve the problem raised by the standard skeptical argument, it does expose its weakness, namely the implicit commitment to standard metaphysics. Therein lies the lesson in the big picture. The skeptical problem, as we typically think of it, is underpinned by a certain metaphysical view of things. And the way the problem is expressed is therefore dependent on whether we accept that metaphysical view. This is similar to the way in which some other antiskeptical strategies show that skeptical arguments depend on a disputable view of perceptual experience.47 Knowing what there is, while being ignorant of whatsk things are, is compatible with knowing very little about the external world, which is the basic skeptical problem. But we should not unreflectively assume the metaphysical view that brings only one, but not the other, into play. Though these general upshots are seldom appreciated in the contemporary literature, they should not be too surprising. Recall what is disturbing about the skeptical problem: that you are ignorant about the world around you, that you don’t know what is going on, that your cognitive life in relation to reality is a joke. It seems clear that your failure to know whether this is all a BIV simulation, or whether this is all a demon’s trick, itself describes just those unfortunate circumstances, whatever else might follow from it about ordinary objects.

47

This is according to some disjunctivist antiskeptical strategies, such as McDowell (1994).

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

References Avnur, Yuval, Brueckner, Anthony, & Buford, Christopher (2011). No Closure on Skepticism. Pacific Philosophical Quarterly 92 (4):439–447. Avnur, Yuval (2012). Closure Reconsidered. Philosophers’ Imprint 12(9). https://quod.lib.umich.edu/p/phimp/3521354.0012?rgn=full+text. Avnur, Yuval (2023). Reality+: Virtual Worlds and the Problems of Philosophy by David J. Chalmers (W. W. Norton & Company, 2022). Philosophy 98 (1): 107–111. Berkeley, George B., & Mathias, Michael B. (2007). George Berkeley: Three Dialogues between Hylas and Philonous. Routledge. Boghossian, Paul A. (1989). Content and self-knowledge. Philosophical Topics 17 (1):5–26. Bouwsma, Oets K. (1949). Descartes’ Evil Genius. Philosophical Review 58 (2): 141–151. Brueckner, Anthony (2010). Charity and Skepticism. In Essays on Skepticism. Oxford University Press. pp. 177–181. Burge, Tyler (1988). Individualism and self-knowledge. Journal of Philosophy 85 (November):649–663. Burnyeat, Myles F. (1982). Idealism and Greek Philosophy: What Descartes Saw and Berkeley Missed. Philosophical Review 91 (1):3–40. Button, Tim (2013). The Limits of Realism. Oxford: Oxford University Press. Chalmers, David J. (2005). The Matrix as Metaphysics. In Christopher Grau (ed.), Philosophers Explore the Matrix. Oxford University Press. pp. 132–176. Chalmers, David (2012). Constructing the World. Oxford: Oxford University Press. Chalmers, David J. (2017). The Virtual and the Real. Disputatio 9 (46): 309–352. Chalmers, David J. (2018). Structuralism as a Response to Skepticism. Journal of Philosophy 115 (12):625–660. Chalmers, David J. (2022). Reality+: Virtual Worlds and the Problems of Philosophy. W. W. Norton. Coliva, Annalisa (2015). Extended Rationality: A Hinge Epistemology. PalgraveMacmillan. Craig, Edward (1990). Davidson and the Sceptic: The Thumbnail Version. Analysis 50 (4):213–214. Dasgupta, Shamik (2015). Inexpressible Ignorance. Philosophical Review 124 (4): 441–480.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

References

51

Davidson, Donald (1986). A Coherence Theory of Truth and Knowledge. In Ernest LePore (ed.), Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson. Blackwell. pp. 307–319. Davidson, Donald (2006). A Coherence Theory of Truth and Knowledge. In The Essential Davidson. Oxford University Press. pp. 225–241. Dretske, Fred I. (1970). Epistemic Operators. Journal of Philosophy 67 (24): 1007–1023. Dretske, Fred I. (2005). The Case against Closure. In Matthias Steup & Ernest Sosa (eds.), Contemporary Debates in Epistemology. Blackwell. pp. 13–25. Goldberg, Sanford C. (ed.) (2015). The Brain in a Vat. Cambridge University Press. Helton, Grace (forthcoming). On Being a Lonely Brain-in-a-Vat: Structuralism, Solipsism, and the Threat from External World Skepticism. Analytic Philosophy. Johnston, Mark (2010). Surviving Death. Princeton University Press. Klein, Peter (1986). Radical Interpretation and Global Scepticism. In Ernest LePore (ed.), Truth and Interpretation. Basil Blackwell. pp. 369–386. Korman, Daniel Z. (2015). Objects: Nothing Out of the Ordinary. Oxford University Press. Kvanvig, Jonathan L. (2003). The Value of Knowledge and the Pursuit of Understanding. Cambridge University Press. Langton, Rae (2004). Elusive Knowledge of Things in Themselves. Australasian Journal of Philosophy 82 (1):129. Leibniz, Gottfried W. (1923). Sämtliche Schriften und Briefe. Akademie-Verlag. 6.4:1502–1503. Lewis, David (2009). Ramseyan Humility. In David Braddon-Mitchell & Robert Nola (eds.), Conceptual Analysis and Philosophical Naturalism. Massachusetts Institute of Technology Press. pp. 203–222. McDowell, John (1994). Mind and World. Harvard University Press. Moore, A. W. (1996). Solipsism and Subjectivity. European Journal of Philosophy 4 (2):220–235. Nagel, Thomas (1986). The View from Nowhere. Oxford University Press. Nozick, Robert (1981). Philosophical Explanations. Harvard University Press. Pappas, George (2008). Berkeley’s Treatment of Skepticism. In John Greco (ed.), The Oxford Handbook of Skepticism. Oxford University Press. Parfit, Derek (1984). Reasons and Persons. Oxford University Press. Parfit, Derek (2011). The Unimportance of Identity. In Shaun Gallagher (ed.), The Oxford Handbook of the Self. Oxford University Press. pp. 13–45. Pritchard, Duncan (2016). Epistemic Angst: Radical Skepticism and the Groundlessness of Our Believing. Princeton University Press.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

52

References

Pryor, James (2005). What’s So Bad about Living in the Matrix? In Christopher Grau (ed.), Philosophers Explore the Matrix. Oxford University Press. pp. 40–49. Putnam, Hilary. (1975). The Meaning of “Meaning.” In Philosophical Papers: Vol. 2. Mind, Language and Reality. Cambridge University Press. pp. 215–271. Putnam, Hilary (1981). Reason, Truth, and History. Cambridge University Press. Putnam, Hilary (1994). Comments and Replies. In P. Clark & B. Hale (eds.), Reading Putnam. Blackwell. pp. 242–289. Roush, Sherrilyn (2010). Closure on Skepticism. Journal of Philosophy 107 (5):243–256. Russell, Bertrand (1927). The Analysis of Matter. Kegan Paul. Schwitzgebel, Eric (2017). 1% Skepticism. Noûs 51 (2):271–290. Sklar, Lawrence (1985). Philosophy and Spacetime Physics. University of California Press. Stroud, Barry (1984). The Significance of Philosophical Scepticism. Oxford University Press. Stroud, Barry (1986). The Physical World. Proceedings of the Aristotelian Society 87 (1):263–277. Sundell, Tim (2016). Eligibility and Ideology in the Vat. In Sanford Goldberg (ed.), The Brain in a Vat. Cambridge University Press. pp. 226–250. Treanor, Nick (2014). Trivial Truths and the Aim of Inquiry. Philosophy and Phenomenological Research 89 (3):552–559. Valberg, J. J. (2007). Dream, Death, and the Self. Princeton University Press. Van Fraassen, Bas. C. (1980). The Scientific Image. New York: Oxford University Press. Vogel, Jonathan (1990). Cartesian Skepticism and Inference to the Best Explanation. Journal of Philosophy 87 (11):658–666. Vogel, Jonathan (2019). Space, Structuralism, and Skepticism. Oxford Studies in Epistemology 6. Walker, Mark (2020). Moore’s Proof, Theory-Ladenness of Perception, and Many Proofs. Philosophical Studies 177 (8):2163–2183. Wright, Crispin (1992). IV – On Putnam’s Proof That We Are Not Brains-in-a-Vat. Proceedings of the Aristotelian Society 92 (1):67–94.

Acknowledgments

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Thanks to David Chalmers, Sinan Dogramaci, Eric Schwitzgebel, and Jonathan Vogel for discussion of these topics. A version of some material in Sections 3 and 4 appears, in shorter form and with a narrower focus, in Avnur, Yuval (forthcoming) “Veridicalism and Skepticism,” The Philosophical Quarterly. Earlier versions of those ideas were presented in Claremont College’s “Work in Progress” seminar and the Skepticism Conference at University of California–Irvine under the title “Unhinged Skepticism,” in 2018 at a University of California–Riverside workshop, and more recently at Central and Pacific APAs. Thank you to all who commented, and to the editor and anonymous referees of this series.

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Epistemology Stephen Hetherington University of New South Wales, Sydney Stephen Hetherington is Professor Emeritus of Philosophy at the University of New South Wales, Sydney. He is the author of numerous books, including Knowledge and the Gettier Problem (Cambridge University Press, 2016) and What Is Epistemology? (Polity, 2019), and is the editor of several others, including Knowledge in Contemporary Epistemology (with Markos Valaris: Bloomsbury, 2019) and What the Ancients Offer to Contemporary Epistemology (with Nicholas D. Smith: Routledge, 2020). He was the Editor-in-Chief of the Australasian Journal of Philosophy from 2013 until 2022.

About the Series

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

This Elements series seeks to cover all aspects of a rapidly evolving field, including emerging and evolving topics such as: fallibilism; knowinghow; self-knowledge; knowledge of morality; knowledge and injustice; formal epistemology; knowledge and religion; scientific knowledge; collective epistemology; applied epistemology; virtue epistemology; wisdom. The series demonstrates the liveliness and diversity of the field, while also pointing to new areas of investigation.

Epistemology Elements in the Series Foundationalism Richard Fumerton The Epistemic Consequences of Paradox Bryan Frances Coherentism Erik J. Olsson The A Priori Without Magic Jared Warren Defining Knowledge Stephen Hetherington Wisdom: A Skill Theory Cheng-hung Tsai Higher-Order Evidence and Calibrationism Ru Ye The Nature and Normativity of Defeat Christoph Kelp

https://doi.org/10.1017/9781009243308 Published online by Cambridge University Press

Philosophy, Bullshit, and Peer Review Neil Levy The Skeptic and the Veridicalist: On the Difference Between Knowing What There Is and Knowing What Things Are Yuval Avnur A full series listing is available at: www.cambridge.org/EEPI