Agnosticism: Explorations in Philosophy and Religious Thought 9780198859123, 9780191891687, 0191891681, 9780192602978, 0192602977

Citation preview

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

A GN O S TI CI S M

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Agnosticism Explorations in Philosophy and Religious Thought

Edited by FRANCIS FALLON and G A V I N H Y MA N

1

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © OUP 2020 The moral rights of the authors have been asserted First Edition published in 2020 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020934748 ISBN 978–0–19–885912–3 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

To an Unknown God

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Contents List of Contributors Introduction Francis Fallon and Gavin Hyman

ix 1

1. The New Agnosticism Robin Le Poidevin

29

2. What is Wrong with Agnostic Belief? Yuval Avnur

47

3. If Agnosticism, Then What? Francis Fallon

81

4. ‘New Agnosticism’, Imaginative Challenge, and Religious Experience David Leech

107

5. Hegel, Wittgenstein, and the Question of Agnosticism Gavin Hyman

140

6. Aquinas and Agnosticism Paul O’Grady

163

7. Redeeming Agnosticism: Richard Kearney’s Anatheistic Wager Guy Collins

186

Index

213

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

List of Contributors Yuval Avnur is Associate Professor of Philosophy at Scripps College, Claremont, USA. He has published numerous articles in journals such as Synthese, Philosophy and Phenomenological Research, Pacific Philosophical Quarterly, Theoria, and International Journal for the Study of Skepticism. Guy Collins is Rector of St Thomas Church and Episcopal Chaplain to Dartmouth College, Hanover, New Hampshire, USA. He is author of Faithful Doubt: The Wisdom of Uncertainty (2014). Francis Fallon is Assistant Professor of Philosophy at St John’s University, New York, USA. He has published in the Review of Philosophy and Psychology and in Topoi, among other places. He is currently involved in empirical work on consciousness as part of Templeton World Charity Foundation’s Accelerating Consciousness Research initiative. Gavin Hyman is Senior Lecturer in the Department of Politics, Philosophy and Religion at the University of Lancaster, UK. His publications include The Predicament of Postmodern Theology (2001), New Directions in Philosophical Theology (ed., 2004), A Short History of Atheism (2010) and Traversing the Middle: Ethics, Politics, Religion (2013). David Leech is Senior Lecturer in the Department of Religion and Theology at the University of Bristol, UK. He is author of The Hammer of the Cartesians: Henry More’s Philosophy of Spirit and the Origins of Modern Atheism (2013) and of articles in journals such as Zygon, Religious Studies and British Journal for the History of Philosophy. Robin Le Poidevin is Professor of Metaphysics in the School of Philosophy, Religion and History of Science at the University of Leeds, UK. His publications include Travels in Four Dimensions: The Enigmas of Space and Time (2003), The Images of Time: An Essay on Temporal Representation (2007), Being: Developments in

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

x

List of Contributors

Contemporary Metaphysics (ed., 2008), and Agnosticism: A Very Short Introduction (2010). Paul O’Grady is Professor in the Department of Philosophy at Trinity College, Dublin, Ireland. His publications include Relativism (2002), Themes in Modern Philosophy (2006), Philosophical Theology (2008), The Consolations of Philosophy: Reflections in an Economic Downturn (ed., 2011), and Aquinas’s Philosophy of Religion (2014).

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

‘For as I went through the city and looked carefully at the objects of your worship, I found among them an altar with the inscription, ‘To an unknown god.’ What therefore you worship as unknown, this I proclaim to you.’ Acts 17:23, New Revised Standard Version *** ‘But why do I thus staggeringly defend myself with one single instance? As if it were not the common privilege of divines to stretch heaven, that is Holy Writ, like a cheverel; and when there are many things in St. Paul that thwart themselves, which yet in their proper place do well enough if there be any credit to be given to St. Jerome that was master of five tongues. Such was that of his at Athens when having casually espied the inscription of that altar, he wrested it into an argument to prove the Christian faith, and leaving out all the other words because they made against him, took notice only of the two last, viz., “To the unknown God”; and those too not without some alteration, for the whole inscription was thus: “To the Gods of Asia, Europe, and Africa; To the unknown and strange Gods.” And according to his example do the sons of the prophets, who, forcing out here and there four or five expressions and if need be corrupting the sense, wrest it to their own purpose; though what goes before and follows after make nothing to the matter in hand, nay, be quite against it. Which yet they do with so happy an impudence that oftentimes the civilians envy them that faculty.’ Erasmus, In Praise of Folly (Gutenberg.org translation)

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction Francis Fallon and Gavin Hyman

Over the last decade or so, debates around atheism have loomed large, both within academic philosophy and theology, as well as in public discussions more generally. The catalyst for these wide ranging debates was the work of the so called ‘New Atheists’, most notably Richard Dawkins, Christopher Hitchens, and Sam Harris. These, in turn, gave rise to numerous replies, predominantly written by Christian theists, many of them theologians, including Alasdair McGrath, Keith Ward, and Ian Markham. The resulting debates have been extended, wide-ranging, and sometimes bad-tempered, taking place across a wide range of media, especially electronic ones, and have certainly not been restricted to university seminar rooms. They have served to revive interest in both religion and atheism in the public sphere, and have revealed that the general public in the Western world is by no means indifferent to these questions. But in some ways, these discussions have been misleading. For as the debates have become increasingly polarized, they have indirectly suggested that people are firmly ensconced within one of two positions, namely theism or atheism. The evidence, however, suggests that this is far from being the case; rudimentary opinion polls as well as sophisticated sociological analyses have suggested that many, perhaps most, in the Western world, feel unable to identify unequivocally with either of these two positions.1

1

For extended discussion of this, see Taylor (2007).

Francis Fallon and Gavin Hyman, Introduction In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0001

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

2

Agnosticism

In light of this, it is perhaps surprising that in the philosophy of religion, there has been so little discussion of the nature, definition, and plausibility of agnosticism.2 Philosophy of religion has long been concerned with evaluating the arguments and evidence in favour of theism, and, more recently, philosophers have given more attention to the arguments and evidence in favour of atheism. Agnosticism, in contrast, has been relatively unexplored, and this in spite of the fact that many in the West would self-identify as ‘agnostic’ or at least would be hesitant about self-identifying as either ‘theist’ or ‘atheist’. One estimate holds that between 450 and 500 million people (7 per cent of the world’s population) are non-believers.3 Here, ‘nonbeliever’ refers to agnostics as well as to those who positively disbelieve in the existence of God.4 This study relies on a survey that defines agnosticism as ‘[not knowing] whether there is a God and [not believing] that there is any way to find out’. Such a reading of agnosticism, as we shall see, is controversial: ‘agnostic’ arguably refers not only to those who believe that there is no way to find out if God exists, but also to those who are open to such demonstration, but do not take themselves to have come across it. Still, even on the narrower definition, the number of agnostics is approximately the same as the number of atheists: well over 200 million people.5 As soon as we consider the issue of agnosticism, we are immediately confronted with a myriad variety of questions, the answers to which are by no means self-evident. How exactly is agnosticism to be defined? Is it an explicit belief or an implied one? Should it be understood as a belief at all, or should it better be viewed as a refusal of all beliefs? Is it mutually exclusive with respect to theism and atheism or can it in some sense be coterminous with one or the other? Is it the espousal of a specifically religious unknowing or of a

2

See Le Poidevin (2010) for a discussion of some of the work that has been done. See Keysar and Navarro-Rivera (2013), p. 554. There are, of course, difficulties in applying European/North American terms such as ‘atheism’ and ‘agnosticism’ globally, particularly in non-Western contexts. 4 The International Social Survey Programme of 2008, which provides one of the most important data-sets on non-belief, does not distinguish at all between positive atheism and agnosticism (Hood, Jr, and Chen (2013), p. 540; for statistics suffering from a similar vagueness—lumping together ‘atheist, agnostic, or non-believer’—see Zuckerman (2007), pp. 55–6). 5 Keysar and Navarro-Rivera (2013), pp. 556–7, 559. 3

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

3

philosophical one as well? Is it simply the outcome of a philosophical impasse or can a philosophical case be made in defence of agnosticism itself? Such questions proliferate further when one contemplates the postmetaphysical turn taken by some in philosophy. To what extent is the conventional taxonomy of theism, agnosticism, and atheism itself the product of a distinctively modern and metaphysical philosophical framework? Is it the case that agnosticism is only really intelligible within such a metaphysical context? Of the three positions, agnosticism seems to be particularly vulnerable to this charge. Wittgenstein and his followers were able to conceive of a non-foundational theism in a post-metaphysical space, as they were able also to envisage a nonfoundational atheism (even if theism and atheism have too frequently manifested themselves in foundational and metaphysical guise). But to what extent can agnosticism be similarly re-conceived? Is agnosticism more wedded to metaphysics than either theism or atheism? In other words, in a post-metaphysical context, is there really anything left for agnosticism to be agnostic about? This sample of questions is illustrative rather than exhaustive, and they indicate the extent to which we enter something of a minefield as soon as the issue of agnosticism is even raised. They also indicate the extent of the inherent interest of the topic, as so much is potentially contested and much classificatory and argumentative work remains to be done. Indeed, such contestation, classification, and argumentation would seem to be required when we recall how many feel unable to identify themselves unequivocally as either theists or atheists. They are all the more needed when we recall the comparative neglect of agnosticism in so many of the popular and public debates, and when, in the academic sphere, we see a plethora of histories, introductions, companions, and handbooks on atheism, but relatively few on agnosticism. In this context, the time is ripe for a thoroughgoing and nuanced investigation into agnosticism, and it is such an investigation that the essays in this volume seek to provide. The contributors of the following essays are united in their common quest to illuminate and interrogate the concept of agnosticism, but they do so from various perspectives. Some are self-confessed agnostics, others are not; some make an explicit philosophical case in defence of agnosticism, while others stop short of this. Some claim to be religious, theistic, or theologians, while others would make no such claims. Some are analytic philosophers, others are continental; some

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

4

Agnosticism

see themselves as contributing to the task of metaphysics, others are seeking a path beyond the metaphysical. The discussions of agnosticism that follow are enhanced rather than encumbered by this confessional and methodological plurality. As differences encounter one another, the complex field of agnosticism is brought to light and the startling variety of what might be considered agnosticism is disclosed. In this process, older understandings of agnosticism are explored and clarified, while newer ones are inaugurated. What this means is that contested definitions of agnosticism are as much at stake as the variously contested confessions, methodologies, and arguments. Several of the essays proffer new understandings of what we might take agnosticism to mean, and thereby suggest new ways in which agnosticism may be defined. What is striking and stimulating about some of these definitions is the way in which they differ from more conventional and intuitive ones; in these instances, we see nothing less than a re-conceptualizing of agnosticism itself. These exploratory and creative innovations do not take place in a vacuum, however; on the contrary, they are situated within a long trajectory of debates about the nature of agnosticism that stretches back to the term’s coining in the nineteenth century. By way of introduction to the essays, therefore, we shall give some attention to the vexed question of how agnosticism has been and may be defined. *** Defining agnosticism proves less simple than one might expect. According to The Oxford English Dictionary (OED),6 the primary meaning of ‘agnostic’ is ‘a person who believes that nothing is known or can be known of immaterial things, especially of the existence or nature of God’. This immediately confronts us with a variety of possible interpretations: The first instance of ‘or’ could be inclusive (nothing is known nor can be known . . . ) or exclusive in one sense (nothing is known, although something might be knowable under the right conditions). Given the second ‘or’ there are six possible delineations of the scope of agnosticism: lack of knowledge—either necessarily or contingently—concerning God’s existence only, concerning God’s nature only, or concerning both God’s existence and nature. The secondary listing describes the extended use (partly) as ‘a person who 6 OED (2017). The 1989 edition and 1971 compact edition also admit various interpretations.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

5

is not persuaded by or committed to a particular point of view’, which could admit the doubly inclusive, and so the widest, delineation. As one would expect of so ambiguous a concept, various subdivisions and distinct forms have been identified, and it will be worth giving some attention to these. First, it has been suggested that agnosticism can itself be divided into a ‘strong agnosticism’ and a ‘weak agnosticism’: ‘Strong agnosticism’ is the conviction that ‘nothing is known or can be known of immaterial things, especially with regard to the existence or nature of God’ (note that this also has the same ambiguities as the OED definition) and ‘weak agnosticism’ is ‘a lack of commitment or conviction’ which, of course, may apply to one’s attitude towards religious belief.7 This distinction echoes, although it is distinct from, Christopher Lane’s own differentiation between ‘permanent’ (strong) and ‘temporary’ (weak) agnosticism.8 ‘Permanent’ agnosticism is effectively agnosticism in principle, which is consequently not open to a change of mind based on further argument, evidence, or a change of circumstances. ‘Temporary’ agnosticism, on the other hand, subscribes to agnosticism for the time being, and is in principle open to a change of mind. Michael Martin elaborates the agnosticism taxonomy by distinguishing between ‘cancellation’ and ‘sceptical’ agnosticism. ‘Cancellation agnosticism’ holds that there are reasons supporting both belief and disbelief in God’s existence (thus cancelling out either conclusion) while ‘sceptical agnosticism’ holds ‘that there are no good reasons for believing that God exists and none for believing that God does not exist’.9 There has also been discussion of whether agnosticism is distinct from atheism, such that the two are mutually exclusive, or whether agnosticism is actually in some sense a species of atheism. The latter view especially comes into view in light of a distinction that is made between positive atheism—the belief that God does not exist—and negative atheism, an absence of belief in God’s existence. Stephen Bullivant, for instance, says that negative atheism ‘includes such positions as agnosticism (in both the classical sense of a specific belief that there is insufficient evidence either to believe or disbelieve in the existence of a God or gods, and in its more popular sense of not having made up one’s mind)’.10 Martin has pointed out that while agnosticism may be inconsistent with positive atheism such that the 7 9

Bullivant and Lee (2016). Martin (2007), p. 3.

10

8 Lane (2012), p. 143, inter alia. Bullivant (2013), p. 14.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

6

Agnosticism

two are mutually exclusive, this is not the case with regard to negative atheism: ‘agnosticism is compatible with negative atheism in that agnosticism entails negative atheism. Since agnostics do not believe in God, they are by definition negative atheists’.11 So on this reading, the family of atheism includes the genera of positive and negative atheism, and the genus negative atheism includes the species agnosticism.12 These associations of agnosticism with atheism should not, however, eclipse the connection between some forms of agnosticism and theism. Indeed, for some, a philosophical agnosticism may be viewed as a purgative preface to religious faith. Kierkegaard, for instance, was indicating something along these lines when he defined faith as ‘an objective uncertainty held fast in an appropriation-process of the most passionate inwardness’.13 For Kierkegaard, the demonstration of an objective fact was of no existential significance, and therefore of no religious interest. Rather, what mattered was the inner appropriation and enactment of the article of faith, about which one might remain ‘objectively’ uncertain. Kierkegaard might thus be in complete accord with the agnostic’s assessment of the reasons, evidence, or arguments for theism, but he would see this as the starting point for religious faith rather than precluding it. For Wittgenstein too, the essence of faith lies precisely in the absence of philosophical evidence, such that if faith sought to ground itself in evidence, it would effectively destroy itself as faith. He says that such arguments for and against faith ‘are, in a way, inconclusive. The point is that if there were evidence, this would in fact destroy the whole business. Anything that I normally call evidence wouldn’t in the slightest influence me’.14 As with Kierkegaard, Wittgenstein would agree with the agnostic that the evidence was inconclusive, but would hold that the agnostic’s error was to seek such evidence in the first place. A quite different way in which agnosticism and theism might be associated is through the incorporation of a certain agnosticism into theistic belief itself, injecting a mystery or unknowing regarding God’s nature and perhaps even God’s existence. In very different ways, this can perhaps be discerned in the thought of both Aquinas and Kant. As Don Cupitt has pointed out, for Aquinas, it might be said that ‘God’s existence is certain but his nature unknowable’, whereas for Kant, ‘it is God’s existence rather than his nature that is 11 13

12 Martin (2007), p. 3. Martin (2007), p. 3. 14 Kierkegaard (1941), p. 182. Wittgenstein (1966), p. 56.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

7

unknowable. God’s nature is not mysterious, for we have a clear, unproblematic and useful idea of God immanent within our reason’.15 So for Aquinas, God’s nature is (in itself) unknowable, whereas for Kant, God’s existence (as such) cannot be an object of theoretical knowledge. But for both of them, these very different agnostic confessions lie within a broader confession of faith. There is space, even apart from the positions of Kierkegaard, Wittgenstein, Aquinas, and Kant, where agnosticism and theism arguably overlap. Of course, many theists believe in God, while at the same time admitting to an uncertainty about God’s existence. Indeed, it may well be that their belief in God is inseparable from that sense of uncertainty. Their belief fits at least some understanding of the term ‘theist’, but their lack of certainty suggests that the term ‘agnostic’ would be appropriate too.16 While some view agnosticism as a sub-species of atheism, and others understand it to be an aspect of theism, it is worth noting that Thomas Huxley, who is credited with coining the term ‘agnosticism’, disavowed both understandings. Huxley, a renowned naturalist, was motivated to coin this new term by his dissatisfaction with the term ‘atheism’, which he believed had come to stand for a position that was too dogmatic. It made a definitive metaphysical claim about the non-existence of God, for which, he thought, there was insufficient evidence. Consequently, for Huxley, atheism was as guilty of metaphysical hubris as was theism. It was in order to avoid what he believed to be these unwarranted instances of metaphysical confidence that he and his associates suggested the term ‘agnosticism’, which betokened not an alternative creed, but a metaphysical unknowing. The term, of course, derives from the Greek word gnosis, knowledge. Thus, it did not denote a denial or negation of theism, but rather a denial of (metaphysical) knowledge, whether of the kind claimed by theism or by atheism. Over the ensuing century, the term caught the imagination of the public, parts of which came to share Huxley’s lack of confidence in either theism or atheism. 15

Cupitt (2002), p. 7. See Holloway (2003), p. 180, as well as Mawson (2013), p. 23, who attributes such a position to Alvin Plantinga. Weatherhead (1990 [1965]) argues that the only way to ‘re-commend’ Christianity to thoughtful people must admit ‘a large degree of agnosticism’ (p. 45); he cites Dostoevsky: ‘My hosanna has come forth from the crucible of doubt’ (p. 361). On ‘theistic agnosticism’ see Lane (2012), p. 143; on ‘agnostic theism’ see Smith (1979), p. 10. 16

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

8

Agnosticism

Huxley’s biographer, Adrian Desmond, goes so far as to suggest that agnosticism was to ‘become the new faith of the West’, and says that the new term had many advantages: it switched the emphasis to the scientific method and its sensual limitations . . . [Huxley] portrayed agnosticism not as a rival ‘creed,’ but as a method of inquiry . . . His was a sect to end all sects: an attempt to clamber on to a higher moral plain, to escape the priests and paupers, Comtists and Christians.17

In spite of his seminal role in the formation of agnosticism, it has been Huxley’s fate to be remembered as much, if not more, for his role as a key protagonist in the drama that came to be emblematic of the supposed Victorian clash between science and religion. This drama was, of course, the debate on evolution between Huxley and Samuel Wilberforce, bishop of Oxford, at a meeting of the British Association in June 1860. Despite its influence on Huxley’s later reputation, the debate gained little attention at first. (The publication of Darwin’s, and then Huxley’s biographies raised its profile.18) Moreover Huxley himself did not subscribe to any necessary clash between science and religion; he agreed with Darwin that the evolutionary hypothesis in no way necessarily entailed atheistic conclusions. This was another reason for his espousal of agnosticism rather than atheism. Darwinism, he believed, was consistent with divine design in that the logical possibility remained that God created the world in a way that set the conditions for the evolutionary processes to follow. Huxley even explicitly stated that the ‘antagonism between science and religion’ was ‘fabricated’ by ‘short-sightedness on both sides’.19 The idea that the natural world might be understood and explained on its own terms without necessarily entailing negative implications for theism has a long pedigree. Albert the Great ‘used the phrase de naturalibus naturaliter to convey the idea that it is legitimate to study nature as if God does not intervene’, and his student Thomas Aquinas, as well as others, followed him in this (Harrison 2010: 12; see also McGinley 2016).20 Indeed, this might be one way of understanding the meaning of ‘naturalism’ in contrast to ‘materialism’.21 The former

17 19 20 21

18 Desmond (1994), p. 375. Chadwick (1970), p. 10. Brooke (2010), p. 109. Emphasis added. Harrison (2010), p. 12; see also McGinley (2016). For further elaboration of this and related distinctions, see Fales (2007).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

9

understands and explains the world on its own terms without prejudging the question of God (or anything transcendent), whereas the latter is positively committed to the denial of God (or anything transcendent). The natural sciences proceed in accordance with the principle of naturalism but are in no way necessarily committed to the principle of materialism. Insofar as naturalism does not pre-judge the question of God, it may be seen to be compatible with atheism, agnosticism, or theism. Harrison has pointed out that the term ‘methodological naturalism’ has come in more modern times to denote the idea that science proceeds (or should proceed) in part by excluding appeal to the supernatural in attempting explanation, where this principle is ‘still consistent with belief on the part of the investigator’.22 This contrasts with metaphysical naturalism, which aligns more with materialism in that it positively denies any supernatural existence.23 ‘Methodological agnosticism’ shares with methodological naturalism a principled avoidance of invoking the supernatural without committing to its denial, putting it at odds not only with metaphysical naturalism, but also with ‘methodological atheism’.24 From Huxley down to the present day, therefore, scientific research and analysis has been seen by practitioners and commentators alike to be perfectly compatible with atheism, agnosticism, or theism. And yet, in the popular mind, the association of ‘science’ with (positive) ‘atheism’ remains strong. Why should this be so? This is not the place, of course, for the development of a complete and thorough answer to this question, although some attempts have been made elsewhere.25 We can observe that this popular preconception has certainly been 22 Harrison (2010), p. 12. Indeed, observing such a consistency sets the tone for The Cambridge Companion to Science and Religion, which announces in its Introduction ‘the idea that religion and science have always been, and are of necessity, ranged against each other is not a view that will find much support in this volume’ (Harrison (2010), p. 16). 23 Harrison (2010); see also Ruse (2010), p. 229. Grayling (Gordon and Wilkinson (2005), p. 4) chooses to call himself a ‘naturalist’ rather than an ‘atheist’, but for different reasons: he believes that using ‘atheist’ (somewhat paradoxically) confers too much legitimacy upon the debate about theism; in other words he declines characterization as atheist just as he would decline characterization as an ‘agoblinest’ (someone who does not believe in goblins): to accept such characterizations concedes more rational plausibility than is merited. For more on this kind of argument, see Garvey (2010) as well as below, this section. 24 Bullivant and Lee (2016). 25 For example, Martin (2007), Harrison (2010), Hyman (2010), Lane (2012), and Bullivant and Ruse (2013).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

10

Agnosticism

reflected and fuelled by the ‘New Atheists’. As we noted at the outset, ‘New Atheism’ is a major factor in the recent growth in interest in ‘unbelief ’ of various stripes. Furthermore, given that at least some of them define themselves not only in relation to theism, but also in relation to agnosticism, we should give some brief consideration to the nature of their stances. The use of the term ‘New Atheism’ itself presents some difficulties. The term ‘is contested, both by those it is intended to describe and by others’, and for good reason.26 For the term, itself a journalistic invention, to merit reception ‘as an analytical category in an academic context’,27 it should refer to something plausibly unitary, and yet, upon analysis, the similarities among its supposed proponents are relatively superficial. Furthermore, their specific arguments for atheism (when they do appear) rarely offer anything particularly new.28 As Thomas Zenk has observed, ‘There is simply no programme or manifesto of “New Atheism” and there is no all-embracing organization, in which all, or even most, of the so-labelled persons are united’.29 Nevertheless, in spite of these difficulties, a consideration of how those most strongly associated with ‘New Atheism’ (Richard Dawkins, Christopher Hitchens, Sam Harris, and Daniel Dennett, sometimes referred to as the ‘Four Horsemen’) treat the notion of agnosticism reveals the importance of more discrete and rigorous investigation. One of the four, Sam Harris, says nothing about agnosticism in his major work on atheism.30 Two of the remaining three don’t give the topic extended discussion, but mention it several times: Hitchens lumps it in with atheism, along with secular humanism and the free-thinkers,31 mentions it in passing in the course of an anecdote,32 and in the context of a brief hypothetical,33 and places it in scare-quotes when attributing it to Darwin.34 Dennett also subsumes

26

27 Bullivant and Lee (2016). Zenk (2013), p. 245. 29 Lane (2012), p. 4; Zenk (2013), p. 256. Zenk (2013), p. 255. 30 31 Harris (2004). Hitchens (2007), pp. 8 and 180. 32 33 Hitchens (2007), p. 33. Hitchens (2007), p. 150. 34 Hitchens (2007), p. 270. Such neglect is not confined to the ‘Four Horsemen’. In the volume Philosophers without Gods: Meditations on Atheism and Secular Life (ed. Louise M. Antony) there are only five mentions of agnosticism, none substantive, save for in Joseph Levine’s cursory treatment: ‘ . . . going agnostic (that great copout) . . . ’ (Levine (2007), p. 29). 28

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

11

agnosticism under the broader category of atheism in his only mentions of the former.35 Dawkins (2008), on the other hand, does give agnosticism as such a hearing. He divides it into Temporary Agnosticism in Practice (TAP) and Permanent Agnosticism in Principle (PAP).36 He asserts that only TAP appropriately acknowledges that the existence of God is a scientific question, and argues that such agnosticism should be very temporary indeed, because we already have the means to assess the probability.37 He rehearses a version of Russell’s ‘teapot argument’.38 In short, if someone posited the existence of a teapot revolving in orbit, undetectable by any scientific measurements, the burden of proof should fall upon the positer, who would fail to satisfy it; the result is that positive belief in the non-existence of the teapot is the only rational position to take. The teapot is conceivable, as are any manner of outlandish things; this does not licence withholding judgements about their non-existence. Their mere logical possibility does not support agnosticism. This, of course, attempts an argument by analogy, where the existence of the teapot stands in for the existence of God. Although this is not the place for an extended critique, we can draw on the earlier treatment of naturalism to help disentangle some of the threads of Dawkins’ argument. Adopting a de naturalibus naturaliter position can accommodate what he calls his ‘central argument’: even Huxley admitted that Darwinism is consistent with divine design, and neither belief in God’s existence nor agnosticism about it obviously precludes optimism about physical explanation of the order of the universe.39 There would exist an irresolvable tension if the teapot analogy succeeds, because this would entail that affirming or maintaining agnosticism about the existence of God would be epistemologically irresponsible on a naturalistic worldview. There are, however, good

35 Dennett (2006), pp. 23, 55, 259, and 300. As has not been uniformly appreciated, Dennett’s book does not purport to argue for atheism, but for a naturalistic investigation into religion. 36 Dawkins (2008), pp. 70–1; this is similar to one of Lane’s distinctions, already mentioned. 37 Michael Martin (2002, 1990) also discusses the question of God’s existence by reference to probability, reaching conclusions similar to Dawkins’. 38 Dawkins (2008), pp. 74–7. 39 For commentary on Dawkins’ argument that includes a contrast with Huxley, see Lane (2012), pp. 163–5.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

12

Agnosticism

reasons to think that the teapot analogy fails.40 Dawkins regards the question of God as a scientific one; presumably, he is willing to admit into a scientific ontology non-observable entities (e.g. electrons) on the grounds of their contributing to successful explanation. Some theists submit that an existing God is part of the best explanation of the universe; agnostics are at least open to the possibility that God exists even if (current) evidence does not permit affirming it as an inference to the best explanation of the universe. Dawkins’ argument for atheism implies either that something else should offer that explanation or that there just is no explanation. But, according to Garvey, if the former is true, then whatever that something else may be, appeal to an existing God as an explanation for it will remain available. On the other hand, Garvey argues, the latter, even if true, is not something we could know to be true: we could only know that we don’t know of any explanation. This does not offer an argument for theism, but it does show that the role of God is plausibly explanatory in a way that an unsubstantiated (but logically possible) teapot in orbit is not. This in turn at the very least causes problems for Dawkins’ positive atheism, indirectly supporting agnosticism. Others have pointed out that Dawkins’ use of Russell’s ‘teapot argument’ reveals the extent to which he regards the existence of God as an empirical question. Insofar as he regards God as a hypothetical object, his understanding of God is actually far removed from what many (perhaps most) theists actually believe. Although theism is notoriously multifaceted and variegated, there are undoubtedly many strands (including the most historically orthodox strands) which would regard a conception of God as an ‘object’, ‘entity’, or ‘person’ as being not only mistaken but actually heretical. Terry Eagleton has been foremost among those making this point. He has said that pre-modern orthodox theology did not see God the Creator as some kind of mega-manufacturer or cosmic chief executive officer, as the Richard Dawkins school of nineteenth-century liberal rationalism tends to imagine—what the theologian Herbert McCabe calls ‘the idolatrous notion of God as a very large and powerful creature’.41

40 The remainder of this paragraph rehearses an argument laid out in much more detail in Garvey (2010). Thanks to Brian Garvey for his input on this passage. 41 Eagleton (2009), p. 6.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

13

For Eagleton, therefore, all of Dawkins’ polemic is directed at a conception of God that is not necessary to theism. His noisy battery of ammunition essentially misses its target. When the noise subsides, it becomes clear how little has been said about agnosticism, and how much remains to be said. While the ‘New Atheists’ may have generated wide public interest around questions of unbelief, their treatments of agnosticism cursorily subsume agnosticism under the broader category of atheism; as we have seen, Dawkins provides the one explicit discussion from this quarter, but— relying on a well-worn and vulnerable argument—fails to appreciate the extent to which his own claims might actually be compatible with agnosticism. They have thus given short shrift to agnosticism on arguably flimsy grounds. The neglect of agnosticism is not unique to the ‘New Atheists’. As we mentioned earlier, in the philosophy of religion generally, there have been comparatively few treatments in the contemporary literature. This neglect looks all the more surprising given that many of the sociological studies also mentioned earlier testify to the significant number of people who align with neither theism nor atheism. There is, furthermore, philosophical interest in a more detailed treatment of agnosticism, given that many of the more interesting and subtle philosophical questions related to belief and unbelief are located there. This is evident from the various essays in this volume, and it is time now to give more detailed consideration to the contents of those essays themselves. *** Although, as we have noted, the chapters in this volume employ contrasting methodologies, and reach distinctive conclusions, there is nonetheless a sense in which the analyses of a number of them converge. Perhaps the most marked convergence lies in the fact that most of the essays argue that agnosticism need not be hostile or negative in its assessment of or relation to theism. Indeed, several of the essays argue that agnosticism may be in some sense compatible with theism or religious belief. This is not an insignificant point, given that, as we have already observed, it is a matter of contention as to whether agnosticism, theism, and atheism are mutually incompatible positions. Does the espousal of one necessarily entail the denial of the others? Historically, more ‘conventional’ understandings of agnosticism have tended to answer in the affirmative. That is to say, they

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

14

Agnosticism

have held that to espouse agnosticism is thereby to reject or deny both theism and atheism. Several of the essays in this volume seek to challenge that assumption. When we look at the concept of agnosticism with more philosophical rigour and nuance, we come to see that such a straightforward incompatibility can no longer be assumed. (This cuts both ways; in at least some of the essays, we see too that agnosticism may not be as incompatible with atheism as is sometimes assumed.) In any discussion or argument of this kind, of course, the definitions of terms being employed are of central importance. It is not surprising, therefore, that in their re-conceiving of the relationship between agnosticism and theism, most of the authors see fit to devote explicit attention to the question of how agnosticism should be defined and understood. Such distinctions and taxonomies play an important role in reaching new conclusions. Indeed, several of the authors speak of what they call ‘New Agnosticism’ in order to distinguish it from older and narrower understandings from the past. As we shall see, our authors do not necessarily agree on what they take this ‘New Agnosticism’ to be. The first four essays share a broad ‘family resemblance’ in that they all reconsider how agnosticism should be defined; in light of these definitions, they assess the relationship of agnosticism to theism or religious belief, and they all broadly conclude that agnosticism may in some sense be compatible with religious belief or religious life. They are also methodologically connected in that they all employ the procedures and assumptions of analytic philosophy. It will be worth considering these essays as a group, therefore, in order to illuminate their common concerns and also to clarify that on which they differ. The convergences and divergences are alike worthy of note. Insofar as the authors of the first four chapters employ the term ‘New Agnosticism’, what exactly do they mean by this? Preliminary to answering this question, Le Poidevin, in Chapter 1, distinguishes three types of agnosticism: truth agnosticism, which is defined in terms of the truth-value of a proposition; existential agnosticism, which is defined in relation to the question of the existence of an entity or group of entities (paradigmatically, the existence of God), and—the least familiar—semantic agnosticism. Semantic agnosticism relates not to the truth-values but the truth-conditions of propositions. As Le Poidevin puts it, ‘Semantic agnosticism is the natural position for those kinds of discourse where we learn to make

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

15

judgements, and so come to appreciate their applicability conditions, before we are introduced to any theory about what grounds them’. In our everyday discourse, there are all kinds of instances where we make judgements and appreciate their applicability conditions without consciously reflecting on what grounds them—Le Poidevin takes modal discourse as being a prime example of this. He says that it might conceivably be applied to moral discourse too—we might fairly confidently engage in conversations about moral truths without necessarily having committed ourselves to an objectivist or projectionist account of morality. But even if we were consciously to reflect on what grounds such judgements, ‘we may still retain our agnosticism, except this time it may be of a more self-conscious and informed kind’.42 Le Poidevin makes the point that subscribing to a semantic agnosticism in this way does not necessarily commit us to a truth agnosticism with regard to the discourse in question. We might, for instance, be agnostic about what philosophically grounds or underpins our moral or modal discourses without thereby becoming agnostic about moral or modal discourses themselves. Le Poidevin suggests that there is a good case for extending semantic agnosticism in this sense to religious discourse. Conventionally, agnosticism in relation to religion has been an existential agnosticism, which brings with it a commitment to truth agnosticism in relation to a whole range of religious propositions. In these forms of agnosticism, a certain semantics is assumed, namely, realism, whereas twentiethcentury philosophy of religion is replete with arguments for a different semantic underpinning for religious discourse, and this allows us to see that the ‘old’ agnosticism was in fact too limited in its scope. When confronted with alternative (realist, non-realist, and fictionalist) accounts of the truth of religion, a semantic agnosticism would entail that we be agnostic about which of these accounts are true. But being agnostic at this (semantic) level would not entail being agnostic about the religious propositions themselves; rather we could commit ourselves to the truth of religious propositions, while being agnostic about the semantic account of the way in which they are true. What Le Poidevin does, therefore, is to advocate a ‘New Agnosticism’, which is presaged by recent developments in the philosophy of

42

Chapter 1, this volume, p. 33.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

16

Agnosticism

religion, and shows how this ‘New Agnosticism’ might actually coexist with religious belief and practice. Yuval Avnur’s analysis in Chapter 2 points in a similar direction, even if his arguments and conclusion are quite distinct. He too draws attention to different types of agnosticism, although not the same ones as those enumerated by Le Poidevin. Avnur suggests that there are two main types of agnosticism, which are important to distinguish: a non-belief (or a suspension of judgement) on a question that might be motivated or determined by any number of factors, and an agnosticism that is determined by a specifically philosophical assessment that gives rise to a position on the insufficiency of evidence. He wants to consider, in particular, the relationship between these two conceptions: does the position that the evidence supports neither belief nor unbelief (agnosticism in the second philosophical sense) necessarily commit one to a suspension of judgement on belief or unbelief (agnosticism in the first sense)? Does one necessarily entail the other? Or, alternatively, may one consistently believe in something one is agnostic about in the second philosophical sense? Avnur here investigates the very possibility of ‘agnostic belief ’, that is to say, the very possibility of believing in something (in God or in a religion, say) that one knows is philosophically unsupported. Would the enactment of such a possibility somehow transgress philosophical rationality? Avnur expresses the question as follows: what is wrong with ‘agnostic belief ’, from an epistemic perspective, which is one that is concerned with accuracy with respect to the evidence? He examines several possible answers to this question, which all take the form: ‘what is wrong with “agnostic belief ” is that it is always (epistemically) irrational’. Avnur shows what is wrong with all the answers to his question that take this form. Having done so, he says that we cannot simply assume that the [second-order] agnostic thinks that we should not, all things considered, believe h. For, in the first place, it is not obvious that second-order agnosticism that belief h is unjustified and, in the second place, even if it did entail that believing h is unjustified, this seems an arbitrary verdict from the perspective of accuracy. The only uncontroversial verdict about belief h is that it is not justified on the basis of evidence, or that it’s not the case that you should believe h.43

43

Chapter 2, this volume, p. 65.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

17

What Avnur suggests, therefore, is that a more careful distinction between different types of agnosticism—a distinction that in the past has been too often overlooked—brings us to see that one is not being epistemically irrational or philosophically unjustified in believing in something about which one is (in philosophical terms) agnostic. We can see, therefore, the close proximity between Le Poidevin and Avnur, as both seek to justify philosophically the bringing together of agnosticism with religious belief. But it is also instructive to note the differences between them. Most obviously, different understandings of agnosticism are at issue. Le Poidevin, we have seen, advocates what he calls a ‘New Agnosticism’, which he defines as specifically a semantic agnosticism, whereas Avnur’s philosophical agnosticism is closer to the more conventional understandings of agnosticism we discussed earlier, but with the critical caveat that Avnur questions whether an acceptance of philosophical agnosticism in relation to a proposition or set of propositions necessarily entails the adoption of an agnostic stance towards the proposition(s) in question, which is what sets him apart from more conventional understandings. There is also a difference between them in the manner of believing that is at stake. For both of them, we might say that the believing in question takes place in the context of a certain agnosticism; still, the manner of believing is not identical. For Le Poidevin, one is agnostic about the semantic status or underpinning of what one believes to be true, whereas Avnur’s account requires no such agnosticism is this sense. One can believe in a ‘fully realist’ way without having to qualify it, provided that one accepts that one’s ‘fully realist’ belief is not one that is philosophically demanded or justified, that the philosophical evaluation leads only to an agnostic conclusion. Avnur argues that there is nothing irrational about such a conjunction. In Chapter 3, Francis Fallon takes yet another approach. Unlike Le Poidevin and Avnur, he does not propose a ‘new’ or qualified definition of agnosticism. He works with a more conventional philosophical or epistemological agnosticism which he takes as a ‘given’, but he is concerned to ask what such an agnosticism would then entail in practical terms, that is to say, in terms of the kind of life that might be lived. Fallon differs from Avnur in that he accepts that a philosophical position of agnosticism towards a proposition or set of propositions should also entail adopting a stance of agnosticism towards those propositions; Fallon does not advocate the possibility of ‘agnostic belief ’ of the kind defended by Avnur. Indeed, he examines some

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

18

Agnosticism

‘epistemologically agnostic positions that nevertheless prescribe belief ’, but says that they ‘ultimately do not achieve what they intend’. The most that could be said, Fallon argues, is that ‘the agnostic position [can be characterized] as one of hopeful ambivalence concerning God’s existence’.44 In this respect, he endorses Gary Gutting’s account, which adopts an open mind towards belief in God’s existence, while also giving good extra-epistemological reasons to hope for God’s existence. He urges the plausibility of the claim that engagement with religious teaching and practice will be conducive to contact with God (should God exist). Respecting the prior claims of science and ethics, he says that one’s best life, including religious engagement within it, will ‘depend on individual inclinations and capacities’. Thus far, Fallon’s account of agnosticism is one that is disposed in a civil way towards religious belief and practice, while stopping short of embracing the possibility of actual religious belief of the kind advocated by Le Poidevin and Avnur. Fallon then proceeds to elaborate what kind of life this might entail in practice. To this end, he undertakes a discussion of the life and work of Oscar Wilde, looking particularly at Wilde’s relationship with religion. At the conclusion of a rich exposition, Fallon says that Wilde’s life suggests the possibility of ‘an openness to religion that may include entertaining religious belief or practice episodically, but without accepting a theistic position as rationally convincing or even as over-archingly supra-rational’.45 He concludes that philosophical agnosticism is actually compatible with a wide range of actions or even lifestyles. At one limit, an agnostic might live a life that is barely distinguishable from that of an atheist, in that such a person ‘prefers not to attend religious rituals or ceremonies, or does so for extrareligious reasons, does not pray, does not reflect on whether God exists, nor discuss it except fleetingly’.46 On the other hand, at the other limit, an agnostic might live a life that is barely distinguishable from a theist. Without an overall acceptance of theism as rational, and thereby remaining an agnostic, ‘this limit case could take the form of a life of regular religious ritual, practice, and reflection, without thereby becoming something other than agnosticism’.47 At this limit, therefore, agnosticism would be compatible with a form of life that is barely distinguishable from that of a practising theist. Thus, 44 46

Chapter 3, this volume, p. 83. Chapter 3, this volume, pp. 111–12.

45

Chapter 3, this volume, p. 98. 47 Chapter 3, this volume, p. 102.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

19

although Fallon does not follow Le Poidevin and Avnur in making a philosophical case for the possibility of what Avnur calls ‘agnostic belief ’, the practical outcome of his position is close to theirs insofar as it allows for the possibility of the practice of what amounts to a religious life within the philosophical parameters of agnosticism. In Chapter 4, David Leech engages with the influential recent work of the philosopher John Schellenberg as way of formulating his own understanding of a ‘New Agnosticism’. This understanding is different from that elaborated by Le Poidevin, however. Schellenberg has invoked the term ‘New Agnosticism’ to distinguish it from the ‘old’ agnosticism which defines itself in relation to a local (Western) understanding of theism. Schellenberg wants to broaden his understanding of agnosticism so that it is defined not (only) in relation to theism but to a perspective he names as ‘ultimism’. ‘Ultimism’ is the religious proposition that ‘there is a reality which is metaphysically, axiologically and soteriologically ultimate’.48 Schellenberg suggests that there may be such an ultimate reality, but that at this relatively early stage in human endeavour, we are not yet in a position to know whether there is. It is in this sense that Schellenberg is an agnostic: ‘he neither believes nor disbelieves in ultimism because he considers that the state of the evidence, at this stage in philosophical enquiry into religion, does not permit him to believe either that it is true or false’.49 But he nonetheless believes that such agnosticism with regard to ultimism also permits a non-doxastic faith in ultimism. Indeed, he believes that there are good reasons for adopting a non-doxastic faith in ultimism as opposed to what he calls a ‘pure scepticism’, or what others might understand as an ‘old style agnosticism’ which refrains from making any kind of faith commitment, non-doxastic or otherwise. This novel proposal has generated wide discussion in recent analytic philosophy of religion. In particular, without taking issue with Schellenberg’s claim that agnosticism is the most appropriate stance towards ‘ultimism’ at the present stage of human enquiry, some have argued that there might be other, more justified, versions of nondoxastic faith than that of ultimism. Two rivals in particular have emerged: a Kantian-inspired ‘religious pluralism’ and a broader and more minimalist religious framework proposition than ‘ultimism’, which James Elliott has termed ‘ietsism’. Leech subjects both

48

Chapter 4, this volume, pp. 107–8.

49

Chapter 4, this volume, p. 109.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

20

Agnosticism

alternatives to close scrutiny, and concludes by defending Schellenberg’s ‘ultimism’ from their attacks. He agrees with Schellenberg that a philosophical agnostic need not rest content with a ‘pure scepticism’ with regard to faith, but that there are good reasons for adopting a non-doxastic faith stance, and, against recent alternatives that have been proposed, argues that ‘ultimism’ remains the most appropriate object for such a faith stance. Leech (following Schellenberg) therefore coverges with Le Poidevin, Avnur, and Fallon in the general direction of their analyses, namely, in arguing that agnosticism is compatible with some form of religious commitment, but his is perhaps the ‘weakest’ or ‘most general’ in terms of the type of commitment that is held to be warranted. Whereas Le Poidevin, Avnur, and Fallon would allow for agnosticism to be combined with some kind of commitment to a particular religious tradition, Leech proscribes any such particular commitment, which he characterizes in a Schellenbergian way as a ‘filled out’ or ‘elaborated’ form of ultimism. He argues that there are no rational grounds for deciding between these elaborated forms of ultimism, and therefore that the non-doxastic faith stance he defends must only be in relation to this generic and open form of ‘ultimism’. We have thus far been noting some of the differences in arguments and conclusions reached by the authors of the first four essays. But we have also noted a broad ‘family resemblance’ between them. Whereas older or more conventional understandings of agnosticism have tended to see the relationship between agnosticism and religious belief as being mutually exclusive, such that agnosticism would proscribe religious belief and practice, the authors of the first four essays all, in different ways, seek to justify some kind of accommodation between philosophical agnosticism and religious belief and/or practice. Insofar as they do this, we might also say that their conclusions draw them somewhat closer to those of some continental philosophers, who likewise see no incompatibility between philosophical agnosticism and religious belief. As we noted towards the beginning of this Introduction, philosophers such as Kierkegaard and Wittgenstein likewise believe the philosophical arguments for theism to be ‘undecidable’ or ‘objectively uncertain’, but for them this in no way precludes the possibility of religious belief; indeed, for both philosophers, such ‘objective uncertainty’ is in many ways a precondition for authentic religious belief. Of course, it must also be acknowledged that our first four authors stop considerably short of embracing the methods or conclusions of

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

21

Kierkegaard and Wittgenstein. They all defend their positions on the basis of arguments derived from the analytic philosophical tradition; none of them defends what might pejoratively be regarded as some kind of ‘fideism’; and they all qualify their defence of religious belief or practice by placing it within a larger context of philosophical agnosticism. Nonetheless, in their conclusions, if not in their methods, they bring analytic philosophical thinking on agnosticism in relation to religion closer to the conclusions of at least some continental philosophers, and insofar as they do, it might be said that they open up vistas for further fruitful conversations between analytic and continental philosophy on the question of agnosticism and perhaps more broadly. We might also say that the remaining three essays undertake a similar transition, albeit from the opposite direction. That is to say, these three essays concern themselves with continental philosophy, theology, or both—modes of thought which have not traditionally been thought to be particularly amenable to or compatible with agnosticism. These essays question that conventional orthodoxy from the other direction. That is to say, they suggest that there is a role for agnosticism even in places where we would not perhaps expect to find it—within the thought of some continental philosophers who are often assumed to shun it, within the thought of key theologians and, indeed, within theology itself. In Chapter 5, Gavin Hyman looks at the work of two continental philosophers, Hegel and Wittgenstein, who would seem to be most hostile to agnosticism as conventionally understood—not just to agnosticism concerning God, but to any kind of epistemological agnosticism. As he says, Hegel is often thought to aspire to ‘Absolute’ knowledge, to be a philosopher with pretensions to ‘know everything’, whose System is supposed to leave nothing out, and to leave nothing ‘unknown’. Wittgenstein, on the other hand, is often thought to repudiate metaphysics in any form, insisting that ‘nothing is hidden’ and ‘all is open to view’; there is nothing, in other words, to be agnostic about.50

While such characterizations are not wholly mistaken, Hyman says that they are misleadingly unnuanced, and he suggests that there is a form of ‘unknowing’, a form of agnosticism, that is intrinsic to the 50

Chapter 5, this volume, p. 143.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

22

Agnosticism

thought of both these heterogeneous philosophers. Undoubtedly, this is a different kind of agnosticism from the more conventional kind produced by modern Western metaphysics, but it is a form of agnosticism nonetheless. Hyman looks at the role played by certain kinds of agnosticism in the thought of Hegel and Wittgenstein, and says that although forms of it are present in the thought of both the earlier and later Wittgenstein, it becomes much less marked in the latter. Although not completely absent, it becomes marginal and much less significant. In contrast, he suggests that ‘knowing’ in Hegel’s thought is always underpinned by a certain ‘unknowing’; that the process of ‘coming to know’ in Hegel’s thought is in a sense enacted by a dynamic interplay of ‘knowing’ and ‘unknowing’. In this sense, a certain kind of ‘New Agnosticism’ may be said to be integral to the very structure of Hegel’s thought. Hyman asks what is at stake in this specific difference between Hegel and Wittgenstein. One way of expressing the difference would be to say that although Hegel believes certain questions to be unanswerable in the sense of our not being capable of formulating them in a single statement, he nonetheless believes it to be heuristically beneficial to ask these unanswerable questions, not least because there is a sense in which they may be answerable, albeit indirectly. Wittgenstein, on the other hand, abolishes the questions together with the answers, seeing them both as products of a misleading metaphysical disease. Drawing on the work of Thomas Nagel, Hyman suggests that the Hegelian approach is the more existentially profitable, speaking as it does to a certain human tendency to ask selfreflexive questions that may not ultimately be answerable, at least not directly. He suggests therefore that Hegel may provide a more fruitful post-metaphysical path than does Wittgenstein, and this because of a ‘New Agnosticism’ that is intrinsic to his thought. While Hyman detects forms of agnosticism in unexpected philosophical quarters, Paul O’Grady in Chapter 6 illuminates the presence of agnosticism in unexpected theological quarters, specifically in the thought of the medieval theologian Thomas Aquinas. This is to raise the question of the compatibility of agnosticism with theology, a question raised—but from the other direction, as it were—by the first four chapters we considered. How might theology conceive of its own relationship to agnosticism and, indeed, of its own coexistence with agnosticism? As we have noted, more conventional forms of agnosticism have understood themselves as having a

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

23

mutually exclusive relationship with theism or theology, and so questions of how agnosticism is to be defined are likely to be central in any attempt to answer this question. In the context of distinguishing between several different understandings of agnosticism, O’Grady draws attention to a particularly pertinent distinction between ‘sceptical agnosticism’ (as exemplified by Anthony Kenny) and ‘apophatic agnosticism’ (as exemplified by Thomas Aquinas). Sceptical agnosticism acknowledges the importance of questions pertaining to theism and atheism, but argues, on philosophical grounds, for the suspension of judgement. Apophatic agnosticism, on the other hand, is a form of agnosticism found within theistic belief ‘which maintains that we do not have knowledge of God, that claims made about God are necessarily inadequate, that in principle no representation can capture God’s essence’.51 O’Grady is concerned to investigate the nature of the relationship between these two types of agnosticism, and does so through an exploration of the work of both Kenny and Aquinas. O’Grady notes that for Kenny, there are two main ways in which one might be led to espouse agnosticism philosophically. One would be to deny the intelligibility of the theist-atheist question, which is, in effect, the position of positivism, and it is not one that Kenny shares. Another group would concede the meaningfulness of the question, but would deny that there are any philosophical grounds available on which to answer it. But Kenny is more specific than this; according to O’Grady, Kenny’s agnosticism rests upon an insight from the theological doctrine of the ineffability of God, that God cannot be captured by human language, that it is quite impossible to speak about God. Such impossibility is what gives rise to Kenny’s position of ‘sceptical agnosticism’. For O’Grady, what is striking about this is that these are the very same grounds which give rise to Aquinas’ very different religious ‘apophatic agnosticism’. How is this to be explained? How can the same underlying insight give rise to two such contrasting renderings of agnosticism—one sceptical and the other theological? O’Grady argues that this difference can be understood in terms of Kenny’s and Aquinas’ contrasting conceptions of being. Following a detailed analysis of these contrasting conceptions, O’Grady concludes that the key difference between them is that

51

Chapter 6, this volume, pp. 166–7.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

24

Agnosticism

Kenny has a bipolar understanding of existence, whereas Aquinas has a scalar understanding. On the bipolar view, one either exists or not at all; no third option is available. Hence the way existence applies to every single existent is the same in each instance—the thinnest possible predicate. Aquinas, on the contrary, thinks of existence in terms of actuality. Each thing’s existence is different in virtue of the degree of actualization present in it.52

This difference has important ramifications. Kenny thinks Aquinas’ account of being is incoherent, as it is incompatible with his bipolar account, which therefore leads him to adopt his ‘sceptical agnosticism’. Others, however, have suggested that although Aquinas’ account of being is different from that of Kenny, it has its own internal coherence and ‘is sufficiently rich to allow for the postulation of the existence of an ultimate reality whose existence is largely unknown by virtue of its superiority to our conceptual capacities’,53 thus opening the way to Aquinas’ ‘apophatic agnosticism’. O’Grady concludes that Aquinas’ views serve to provide an important challenge, both to ‘insouciant agnostics’ and also to theists who are overly confident of their grasp of the deity. This last point is one taken up and greatly developed by Guy Collins in the final chapter of this volume, in that he too invokes a form of agnosticism that seeks not to negate or to deny theism, but which, operating on and within the matrix of theology itself, puts into question any overly confident grasp of the deity. As we have seen to be the case in several of the other essays, Collins again explores the possibilities for a ‘new agnosticism’ that is not mutually exclusive in its relationship to theism, but which may in a certain sense co-exist with it. Collins suggests that ‘agnosticism may be viewed as a phase within both theism and atheism that opens them up to one another in fruitful dialogue. Indeed, from the perspective of Christian theology, agnosticism is a necessary phase within Christian belief ’.54 He clarifies that he understands by ‘phase’ the scientific meaning of the term as a relationship in time between something that oscillates or repeats, such as an alternating electrical current . . . it may well be that agnosticism can be likened to a continual phase of oscillation between theism and atheism that presents some essential truth about the divine.55

52 54

Chapter 6, this volume, pp. 180–1. Chapter 7, this volume, p. 190.

55

53 Chapter 6, this volume, p. 181. Chapter 7, this volume, p. 190.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

25

In other words, agnosticism may be understood as something that operates upon and within theism, and which prevents it from congealing into a static or settled theism. To develop this idea, Collins explores the work of the continental philosopher and literary critic Richard Kearney, who has coined the term ‘anatheism’ to designate a territory of thought that lies outside dogmatic theism and militant atheism. In this sense, it occupies terrain traditionally associated with agnosticism. But whereas agnosticism has traditionally reserved judgement on the question of theism versus atheism, Kearney’s anatheism does something quite different; beyond the certainties of theism and atheism, it wagers on a faith beyond faith, on what remains to be said about God when the substantial God of metaphysics has been left behind. Collins clarifies what Kearney means by a wager of this kind: ‘Wagers are orientations to the future, not settled positions. The tentativeness of anatheism is part and parcel of its identity . . . In this sense, anatheism describes an oscillating movement that connects some of the most traditional insights of theism with some of the deepest held beliefs of atheism. Kearney cannot be certain about the anatheistic God. Instead, the anatheistic wager incorporates a deep-seated unknowing about God’.56 In the body of his chapter, Collins fleshes out what such ‘anatheism’ entails in practice, showing how Kearney draws on diverse resources from continental philosophy, literature, and traditional theology. On Collins’ reading, Kearney’s ‘anatheism’ occupies the territory that was once inhabited by more conventional forms of agnosticism, but re-imagines it. He says that for Kearney, the possibility of God is strictly undecidable. For him God is possibility above all. And with possibility to be there is also the possibility not to be. If anatheism came down hard on one side of the possibility of God rather than the other it would not be entirely unjust to view it alongside other theisms or atheisms. But as there is an undecidability at its heart that seeks to return to the possibility of God after both theism and atheism, it seems prudent to view anatheism as instead a re-imagined experience of agnosticism.57

56 57

Chapter 7, this volume, p. 193. Chapter 7, this volume, p. 210. Our emphasis.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

26

Agnosticism

We are here in very different terrain, both in methodology and content, from the chapters with which we began. As we engage with the various chapters in this volume, we encounter some diverse methodology and content. But what is particularly striking is that in the midst of these heterogeneous studies, we can glimpse the emergence of some kind of convergence. The earlier chapters, working from within the mainstream of analytic philosophy, and beginning with received understandings of agnosticism, explored and developed these to ask whether agnosticism might not after all be compatible, in different senses and to varying degrees, with religion. The later chapters, engaging more with continental philosophy, and beginning from the standpoints of continental philosophers and theologians, asked whether such thinkers might not after all be compatible, in different senses and to varying degrees, with agnosticism. Undoubtedly, we should not downplay the differences, both in methodology and content, between these various chapters. Indeed, we have been concerned to highlight some of these differences, which are themselves instructive, in this Introduction. But it is equally instructive to note a certain convergence that these various essays point towards—a convergence between agnosticism and religion, and between analytic and continental approaches. This convergence does not serve to reach one common conclusion, but it may serve to prepare the way for further fruitful conversations on the relationship between agnosticism and faith, and between analytic and continental philosophy.

BI B LI OGR APHY Antony, Louise (ed.). Philosophers without Gods: Meditations on Atheism and Secular Life (Oxford: Oxford University Press, 2007). Brooke, John Hedley. ‘Science and Secularization’. In The Cambridge Companion to Science and Religion, edited by Peter Harrison (New York: Cambridge University Press, 2010), pp. 103–23. Bullivant, Stephen. ‘Defining “Atheism” ’. In The Oxford Handbook of Atheism, edited by Stephen Bullivant and Michael Ruse (New York: Oxford University Press, 2013), pp. 11–21. Bullivant, Stephen, and Lois Lee. A Dictionary of Atheism (Online http:// www.oxfordreference.com.jerome.stjohns.edu:81/view/10.1093/acref/ 9780191816819.001.0001/acref-9780191816819-e-1?rskey=wLfkV9& result=1 (Oxford: Oxford University Press, 2016).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Introduction

27

Bullivant, Stephen, and Michael Ruse (eds.). The Oxford Handbook of Atheism (New York: Oxford University Press, 2013). Chadwick, Owen. The Victorian Church, Vol. II (London: A & C Black, 1970). Cupitt, Don. Is Nothing Sacred? The Non-Realist Philosophy of Religion: Selected Essays (New York: Fordham University Press, 2002). Dawkins, Richard. The God Delusion (New York: Mariner Books, 2008). Dennett, Daniel. Breaking the Spell: Religion as a Natural Phenomenon (New York: Viking, 2006). Desmond, Adrian. Huxley: The Devil’s Disciple (London: Michael Joseph, 1994). Eagleton, Terry. Reason, Faith and Revolution: Reflections on the God Debate (New Haven: Yale University Press, 2009). Fales, Evan. ‘Naturalism and Physicalism’. In The Cambridge Companion to Atheism, edited by Michael Martin (New York: Cambridge University Press, 2007), chapter 7. Garvey, Brian. ‘Absence of Evidence, Evidence of Absence, and the Atheist’s Teapot’. Ars Disputandi 10 (2010), pp. 9–22. Gordon, Mick, and Chris Wilkinson (eds.). Conversations on Religion (London: Continuum, 2005). Harris, Sam. The End of Faith: Religion, Terror, and the Future of Reason (New York: Norton, 2004). Harrison, Peter (ed.). The Cambridge Companion to Science and Religion (New York: Cambridge University Press, 2010). Harrison, Peter. ‘Introduction’. In The Cambridge Companion to Science and Religion, edited by Peter Harrison (New York: Cambridge University Press, 2010), pp. 1–17. Hitchens, Christopher. God is Not Great: How Religion Poisons Everything (New York: Twelve, 2007). Holloway, M.R. ‘Agnosticism’. In New Catholic Encyclopedia, edited by Berard L. Marthaler (New York: Thomson-Gale, 2003), pp. 180–4. Hood, Jr, Ralph, and Zhuo Chen ‘Conversion and Deconversion’. In The Oxford Handbook of Atheism, edited by Stephen Bullivant and Michael Ruse (New York: Oxford University Press, 2013), pp. 535–49. Hyman, Gavin. A Short History of Atheism (London: I. B. Tauris, 2010). Keysar, Ariela, and Juhem Navarro-Rivera, ‘A World of Atheism: Global Demographics’. In The Oxford Handbook of Atheism, edited by Stephen Bullivant and Michael Ruse (New York: Oxford University Press, 2013), pp. 553–86. Kierkegaard, Søren. Concluding Unscientific Postscript, tr. David F. Swenson and Walter Lowrie (Princeton: Princeton University Press, 1941). Lane, Christopher. The Age of Doubt: Tracing the Roots of Our Religious Uncertainty (New Haven: Yale University Press, 2012).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

28

Agnosticism

Le Poidevin, Robin. Agnosticism: A Very Short Introduction (New York: Oxford University Press, 2010). Levine, Joseph. ‘From Yeshiva Bochur to Secular Humanist’. In Philosophers without Gods: Meditations on Atheism and Secular Life, edited by Louise M. Antony (Oxford: Oxford University Press, 2007), chapter 2. Martin, Michael. Atheism: A Philosophical Justification (Philadelphia: Temple University Press, 1990). Martin, Michael. ‘Should Agnostics Be Atheists?’ The Philosophers’ Magazine, Summer (2002), pp. 17–19. Martin, Michael (ed.). The Cambridge Companion to Atheism (New York: Cambridge University Press, 2007). Martin, Michael. ‘General Introduction’. In The Cambridge Companion to Atheism, edited by Michael Martin (New York: Cambridge University Press, 2007), pp. 1–7. Martin, Michael. ‘Atheism and religion’. In The Cambridge Companion to Atheism, edited by Michael Martin (New York: Cambridge University Press, 2007), pp. 217–32. Mawson, T.J. ‘The Case Against Atheism’. In The Oxford Handbook of Atheism, edited by Stephen Bullivant and Michael Ruse (New York: Oxford University Press, 2013), pp. 22–37. McGinley, Donall. ‘Can Science and Religion Meet over their SubjectMatter? Some Thoughts on Thirteenth and Fourteenth Century Discussions’. In Robert Grosseteste and the Pursuit of Religious and Scientific Learning in the Middle Ages, edited by Jack P. Cunningham and Mark Hocknull. Studies in the History and Philosophy of Mind, Vol. 18 (Switzerland: Springer, 2016), pp. 263–80. OED (Oxford English Dictionary) (Oxford: Oxford University Press, 2017). Ruse, Michael. ‘Atheism, Naturalism, and Science: Three in One’. In The Cambridge Companion to Science and Religion, edited by Peter Harrison (New York: Cambridge University Press, 2010), pp. 229–43. Smith, George H. Atheism: The Case against God (New York: Prometheus Books, 1979). Taylor, Charles. A Secular Age (Cambridge, MA: The Belknap Press of Harvard University Press, 2007). Weatherhead, Leslie. The Christian Agnostic (New York: Abingdon Press, 1990 [1965]). Wittgenstein, Ludwig. Lectures and Conversations on Aesthetics, Psychology and Religious Belief, edited by Cyril Barrett (Oxford: Basil Blackwell, 1966). Zenk, Thomas. ‘New Atheism’. In The Oxford Handbook of Atheism, edited by Stephen Bullivant and Michael Ruse (New York: Oxford University Press, 2013), pp. 245–60. Zuckerman, Phil. ‘Atheism: Contemporary Numbers and Patterns’. In The Cambridge Companion to Atheism, edited by Michael Martin (New York: Cambridge University Press, 2007), pp. 47–65.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

1 The New Agnosticism Robin Le Poidevin

I. THE ROCKET AT THE BOTTOM OF THE GARDEN Sometime in the summer of 1968, a friend and I decided to convert the old wooden summerhouse at the bottom of my parents’ garden into a rocket. (This was, recall, a year before the first Moon landing, and just seven years after the first manned spaceflight. Britain was lagging behind in the space race, so this was something of a pioneering venture.) The process of transformation, it has to be admitted, was largely cosmetic. Matters of propellant, air-tightness, navigation, and so on, were postponed while we tackled the vital issue of how to make it look like a rocket—from the inside at least. Outside, it looked pretty much the same as it did before, so any Russian or American spies would have suspected nothing. To this aesthetic end we imported various bits of electronic apparatus, long abandoned to a corner of the attic: radios, a television set (with valves which glowed in a satisfying way when it was plugged in), a profusion of wires, an old transformer, and so on. Our preparations did occasionally turn to the practical: how were extra air supplies to be stored? My friend had the idea of trapping the air in bubbles in large jars of water (specifically, the massive jars my grandmother used to store fruit in over the winter), to be released when needed. That the jars would actually contain less air this way than they would if they were simply empty did not trouble us. We were very young, let me add in self-defence, but what exactly was going on? It’s tempting, of course, to say that this was simply and Robin Le Poidevin, The New Agnosticism In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0002

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

30

Agnosticism

paradigmatically a children’s game of make-believe. We didn’t really think that we were building a rocket that could actually take off. We were simply pretending that it was, or would in due course become, a rocket. That was why the look was so important, and the practical obstacles either brushed aside, or dealt with fairly perfunctorily. After all, if this isn’t the right description of our state of mind, then we seem to be left with the uncomfortable alternative that here were two sadly, if forgivably, deluded children, soon to be crushingly disabused by the inevitable failure of their enterprise. Both of these rival descriptions—that we were pretending, and that we were deluded—strike me as a little too neat. Against the pure make-believe hypothesis is the fact that we were prepared, at least sometimes, to come face to face with the issue of air supply (even if our solution was risible). The problem of how to protect what was after all a wooden structure from the intense heat generated by reentry into the Earth’s atmosphere was one that was admittedly beyond our conception, but we certainly considered the issue of propulsion. Here a ready method suggested itself: having seen fireworks in action, we reasoned that attaching a sufficient number of 6d1 rockets to the outside of the summer house would, once they were ignited in a cunning sequence, allow us to resist the forces of gravity. It was not, then, just a matter of appearances. But against the delusion hypothesis is the undoubted fact that at no point in our preparations were we paralysed by fear at the prospect of the awful dangers we faced. If it was just pretence, then of course no danger existed. But if neither make-believe nor delusion, then what? Could it have been an amalgam of these two? This doesn’t seem to be very promising. After all, one state of mind involves believing that we were building a rocket, and the other involves believing that we were not (really) doing so. Perhaps just this conflict of attitudes is realized in certain cases of self-deception, but were we self-deluded? I would say not. Perhaps our attitude could have been described as an oscillating one. We would certainly have resented any remarks from sceptical siblings that the project could not possibly succeed—an observation that would have been quite irrelevant had it simply been a game of make-believe. (Though one is, admittedly, similarly unappreciative of the comment that ‘it isn’t real’ when immersed in a fiction.) On the 1 That is, 6 pennies in the British pre-decimal currency of the time. Equivalent to 2½p today, or (roughly) 3 US cents.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

31

other hand, we would probably have reassured anxious parents, had they taken us as being entirely in earnest, that we were ‘only pretending’. Still, the fact that we would have said this doesn’t entail that that would precisely have captured what we thought. Clearly, we were aiming at some verisimilitude in decorating the inside of the summer house to represent what we imagined the controls of a rocket to look like. But to what end? To aid the fantasy, is the obvious suggestion. The various bits of (inert) gadgetry were simply props in our game. But we were also, at that age, credulous, ready to believe in magic, supernatural forces, ghosts, and so on. It would not have been too much of a stretch, against this background of epistemic dispositions, to believe that a suitable combination of ingredients could, by an entirely natural process, take on the kinds of powers we were, in our imaginations, invoking. We did not expect to understand, in any detail, the mechanism behind devices such as radios and televisions, yet we saw that something that looked a certain way inside (valves, dials, wire coiled around cylinders, etc.) could lead to remarkable effects. Perhaps something would therefore emerge from the grotesque structures we were assembling. I have no reason to suppose we were anything other than ordinary children, with ordinary capacities to understand, and ordinary propensities to play. And I want to take care not to over-intellectualize our thoughts. In fact, it is that desire not to over-intellectualize them that leads me to suppose that there could have been an element of indeterminacy about those thoughts. I don’t suppose we fully believed that this ramshackle wooden shed was going to leap into the stratosphere and beyond. But I do think we imagined that something— quite what, we didn’t know—might just happen. We might become aware of some force, perhaps no more than a distant shadow of those involved in rocketry. Or perhaps we would pick up strange signals on our Heath Robinson set-up, which would put us in touch with the world of space. At least (I am inclined to say) there was a component to our excited imaginings that transcended the pretence that no doubt formed part of our activities. Memory is notoriously unreliable, of course, but whether or not my record of how I spent part of a summer half a century ago is accurate is perhaps not so very important. What I have no more than sketched so far is, I think, a possible state of mind, one that is not determinately and wholly one of belief, nor determinately and wholly one of makebelieve, but perhaps something between the two. I am inclined to

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

32

Agnosticism

think that it is a kind of agnosticism, and one that might provide a useful, even novel, model for religious thought and practice. So now let us leave, for the time being, this childhood vignette, and try to introduce a bit more precision into our discussion.

II. THREE KINDS OF AGNOSTICISM Agnosticism, of any kind, involves not knowing something. This might be a provisional position, while further data or arguments are being gathered or assessed, or it might be a principled one, to the effect that knowledge of a certain kind is not to be had. The term is used in a variety of contexts, not just religious ones. It might be applied to science (concerning the status of theoretical entities, for example), or morality, or economics, and so on. And we might classify agnosticisms by their subject matter (quarks, ghosts, the advantages of a single European currency, and so on). But, whatever the subject matter, we can usefully identify three kinds of agnosticism, on the basis of the kind of object to which they are directed. Truth agnosticism is defined in terms of the truth-value of a certain proposition. One might, for instance, profess agnosticism over the result of a general election. That is, you do not know (and perhaps could not know) whether it is true or false that the Liberal Democrats/ Labour Party/Tories/Whigs will gain a majority. And one might be truth agnostic about a whole domain of discourse, such as any proposition concerning the fundamental structure of matter, or time prior to the Big Bang. Existential agnosticism is concerned with the existence of a certain entity or group of entities. Standard agnosticism over the existence of God would be the paradigm example of existential agnosticism. Equally, one might be agnostic over the existence of a most fundamental particle (perhaps matter is infinitely divisible into yet more fundamental particles, without limit). Existential agnosticism could be represented as a sub-category within truth agnosticism, since if one is agnostic over the existence of Fs, one is also agnostic over the truthvalue of the proposition ‘Fs exist’ or perhaps any proposition predicating intrinsic properties to Fs. Semantic agnosticism is the least familiar. It concerns, not the truth-value, but the truth-conditions of propositions of a given kind.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

33

Here is a plausible example. Modal discourse consists of propositions concerning the possible, the necessary, the impossible, and the contingent, including counterfactual discourse about what else would have been the case had such and such been the case. We show quite a high level of confidence in making everyday modal judgements: that it is not possible to attend a certain meeting, that we would have missed the show if it had not been for a lift from a friend. And a similar level of confidence in some of our less everyday modal judgements, such as that that the square root of 144 is necessarily 12, that it is possible (in a very broad sense) for there to be a golden mountain, but not possible for there to be a round square, and so on. We may reasonably require of any candidate semantics for these modal judgements—that is, a theory which offers an account of their truth-conditions—that it preserve to a large extent the truth of these pre-theoretical judgements as to what is or is not possible. But as to which semantics is the right one, we may remain agnostic. That is, we may be uncertain whether the truth-conditions involve really existing entities known as ‘possible worlds’ or not, or whether such worlds are best thought of as abstract entities on a par with numbers, or entities that are just as concrete (in the sense of having spatial and temporal properties) as our world.2 Semantic agnosticism is the natural position for those kinds of discourse where we learn to make judgements, and so come to appreciate their applicability conditions, before we are introduced to any theory about what grounds them. This, I take it, is exactly the position we are in with respect to modal discourse. But even after we have been introduced to such a theory, we may still retain our agnosticism, except this time it may be of a more self-conscious and informed kind. The case of moral judgements provides another possible example of semantic agnosticism, but one that is perhaps not entirely parallel to the modal case. Here, as with modality, we are introduced to an area of discourse without an accompanying theory as to what grounds it. As we might put it, we learn to make first-order judgements (‘Cloning is permissible’) before we encounter, or come to any conclusions concerning, a second-order account of what grounds 2 Contrast the modal agnosticism presented in Divers (2006), according to which a concrete possible worlds account of the truth-conditions of modal judgements is committed to, but we remain agnostic about whether there are any such worlds.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

34

Agnosticism

those judgements in terms (e.g. an account according to which the statement that ‘Cloning is permissible’ is made true by objective, mind-independent moral values, or one in which the relevant truthconditions are simply our own affective responses). We learn to judge murder as bad, and helping others as good, at a fairly early stage of theoretical development. So we might suppose that our first-order judgements do not reflect any second-order view. Against this, however, John Mackie argues that our ordinary moral judgements have an objectivist assumption at their heart: that moral values attach to actions independently of our judgements or reactions. He thinks this assumption mistaken, and that the correct account of moral values is a subjectivist one, but this he presents as an error theory of moral judgement, not a description of our ordinary understanding.3 If this is right, then we are not ab initio semantic agnostics about moral judgement. Once, however, we encounter the arguments for and against objectivism, and for and against the varieties of subjectivism, and find that we cannot, for the time being, decide which is the correct account, then we become semantic agnostics with respect to moral discourse. So what is the relation between semantic agnosticism and the other kinds, truth and existential? It depends on the case, and on the range of our semantic agnosticism. I suggested above that semantic agnosticism about modality need not be accompanied by truth agnosticism about modal judgements: we can go on confidently asserting that it was possible for any object to be larger or smaller than it actually is, without having first established that there really are possible worlds, or that possibilities are ideas in the mind of God, or practically any other theory. This would, however, presuppose a stance on whether modal sentences were propositional or not. For if they were not propositional, then they would lack truth-values. So any agnosticism as to whether they were propositional would entail truth agnosticism: they might be true, or they might be neither true nor false. A similar situation exists with respect to moral judgements. If our agnosticism about moral semantics is not too extensive, we can go on treating our moral judgements as (by and large) true, without first taking a stance on whether the right account of those judgements was a realist or a

3

See Mackie (1977), pp. 30–5.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

35

projectivist one. On the other hand, if we felt we could not rule out an expressivism about moral discourse, on which moral judgements are not propositional at all, but simply express feelings or commitments, then we would have to remain truth agnostics about moral utterances. There will, then, be a variety of connections between the three kinds of agnosticism just outlined, for any given subject matter, and any given background of assumptions, but the point is that they are different positions, and can in certain circumstances, be detached from each other. So, from science, modality, and morals, let us now turn to theological discourse.

III. IS THIS A GAME? DEFINING RELIGIOUS SEMANTIC AGNOSTICISM In theological matters, the standard kind of agnosticism is most appropriately classified as existential or truth agnosticism: the standard agnostic does not know, professes not to know, and perhaps even asserts the impossibility of knowing, whether or not God exists. This is existential agnosticism. And with it comes truth agnosticism over a whole range of apparently propositional thoughts and sentences, of varying levels of generality: God exists, God created the world, God loves us, God has a plan for us, God (as Christ) died for our sins, and so on. Standard agnosticism about God is not, however, a form of semantic agnosticism. A certain semantics is presupposed, and it is a realist one: that is, the truth-conditions are conceived in terms of the existence and properties of a transcendent god, and that god’s relations to us and the cosmos. Whether theological propositions are true or false depend, not on our judgement, but on a reality quite independent of our judgements. What is a matter of doubt, in standard agnosticism, is whether the truth-conditions obtain. But there is room, I think, for semantic agnosticism in theological matters. Since the mid-twentieth century, a number of positions have emerged in philosophy of religion articulating various alternatives to this realist conception of theological discourse, rejecting one or other of realism’s assumptions, namely: • That theological statements are truth-apt (capable of being true or false);

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

36

Agnosticism • That the meaning of such statements can be read off from a more or less literal understanding of the terms involved; • That their truth-conditions involve states of affairs which are independent of human thought or responses.

Rejecting the first of these features of realism, religious expressivism holds that theological discourse is non-propositional or fact-stating, and expresses religious feelings, experiences, and ideals, and/or a commitment to certain kinds of action (to a life of virtue, for example). And of course, dispensing with this aspect of realism entails dispensing with the others too. Rejecting the second (but not the first), religious positivism holds that the true meaning of theological discourse does not coincide with its apparent or surface meaning, but is to be explained in terms of moral truths or ideals. Whether that entails rejecting the third depends on the view one takes on the truthconditions of moral discourse. Finally, retaining the first two, but rejecting the third, what I shall call religious fictionalism holds that theological statements are true in the relevant religious fiction, but have no application outside that fiction. Thus it is true that God created the world in the same sense that Lear divided his kingdom between his two elder daughters, cursing the youngest and (until that point) most beloved. Does this involve a revision in our understanding of the meaning of theological statements? It may involve a shift in our understanding of their reference, but not, surely, of their meaning. In reading a fiction, we take the meanings of words to be the same as they would be in a work of non-fiction. If this were not so, fiction would be completely baffling.4 Now, without yet judging the viability of realism, one might take a view as to which of these non-realist positions was the most promising. Against expressivism we can urge two considerations. First, that theological sentences certainly appear to have a propositional structure. That is, they look like the kind of thing that we use to state facts, 4

An expressivist account seems to be hinted at in some of Wittgenstein’s writings on religion (see Wittgenstein (1966)), and less ambiguously in the work of D.Z. Phillips ((1976), see, e.g., p. 181). Braithwaite (1971 [1955]) combines expressivist and fictionalist elements. Bultmann’s remark ‘I am interpreting theological affirmations as assertions about human life’ (Bultmann (1972), p. 107) is an explicit affirmation of positivism. And all three are apparent in Cupitt (1980). See Eshleman (2005) for a careful development of a fictionalist account, though it stops short of the fictionalist semantics offered here. Two kinds of fictionalism are discussed in Le Poidevin (2015), with an emphasis on the one which offers a non-realist semantics.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

37

and that in using them we aim at truth. Second, it is not at all clear that expressivism can give a sufficiently fine-grained account of the meaning of different theological sentences. If they are not fact-stating, then their meaning will have to be spelt out in other terms, such as the religious feelings they express. But it is unlikely that there could be a one-to-one correspondence between sentence-type and associated feeling. A variety of sentences, apparently quite different in content, could correspond to one and the same religious feeling. Important differences are lost. So expressivism faces what we might call the problem of loss of content. Positivism faces a similar problem. Although positivism concedes that theological sentences are (or can be) propositional and factstating, the facts they state, it avers, are moral facts (and perhaps, in addition, spiritual ones, insofar as this is not simply a subset of the moral). But will there be a one-to-one correspondence between theological statement and moral fact? No, again there will be a many-one relationship, where different statements correspond to the same moral truth. So positivism too faces the problem of loss of content. In contrast, fictionalism does not propose a reconstruction or revision of the meaning of theological sentences: they can be taken at face value, at least as far as meaning, though not reference, is concerned. So there will be no worries that a fictionalist construal will not be fine-grained enough: it will be exactly as fine-grained as realism. Richard Braithwaite, for whom religious utterances have the function of expressing commitment to a way of life (a version of the expressivist position defined above), answers the objection that this does not distinguish the meaning of different religious utterances by adopting a holistic approach: the meaning of an utterance cannot be given when taken in isolation; rather, it has a meaning by virtue of its relation to the discourse as a whole. So it is the discourse as a whole which is characterized by the commitment to a way of life. The question that remains is how the relations between the different components of the discourse, which define those components’ meanings, are themselves to be defined. And at that point we may need to look beyond expressivism. The points here generalize to positivism. Braithwaite raises another problem for his own expressivism. Different religions appear to correspond to rather similar codes of conduct (setting aside differences in liturgy), which on the expressivist view would collapse them into a single religion. So how are they

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

38

Agnosticism

to be differentiated? It is at this point that Braithwaite brings in what looks like a distinctly fictionalist element: The really important difference, I think, is to be found in the fact that the intentions to pursue the behaviour policies, which may be the same for different religions, are associated with thinking of different stories (or sets of stories). (Braithwaite (1971 [1955]), p. 84)

But having got this far, it is tempting to go back and use the fictionalist element to define the differences in meaning between different religious statements. So much for the question of meaning. What of the religious significance of these positions? In assessing this, we may appeal to a number of functions religions could have (this is not intended to be an exhaustive or definitive list): i) to contribute to cultural identity; ii) to foster certain moral commitments; iii) to evoke and reinforce certain emotions, as part of the mechanism behind (ii); iv) to diminish one’s sense of alienation in the world. As Braithwaite’s suggestions referred to above make clear, fictionalism offers a more ready means than does expressivism or positivism of providing (i), insofar as it is in large part the religious narratives that distinguish different religious cultures. With respect to (ii), in contrast, we may think that the non-fictionalist positions have the advantage, in that they make more explicit the connection with moral commitment. The fictionalist, however, can appeal to the alreadyrecognized phenomenon of the morally transformative effect of stories (paradigm examples being the parables of Jesus), via the effect on our emotions, as well as our judgement. It thus offers a ready means to effect (iii). (iv) is trickier—and indeed here the realist may point out that none of the non-realist positions comes close to achieving the success of the realist solution to this. Reflecting on these kinds of comparison at rather greater length than this very brief summary suggests, one may come to conclusion that, of the non-realist positions, fictionalism is the one to be preferred. However, when we come to what would now be a two-horse race between fictionalism and realism, we may feel at a loss. To be clear, what is at issue here is not the descriptive one of how people in fact use theological discourse. It may be very evident that a realist

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

39

semantics is intended, that when certain people talk of God, their intention is to refer to a transcendent being. And it may equally be evident that others are not employing the discourse in this way. The question, rather, is a normative one: how should we understand theological discourse? What understanding best captures the importance it has for a religious life? And here the answer may not be clear. Non-realists may urge (have, indeed, urged) that a realist semantics makes theological discourse appear to be a kind of scientific theory of humankind and the universe, and a not-very-well-confirmed one at that. Thus understood, it neither offers a satisfactory explanation of observed phenomena, nor does it have any predictive power. Realism thus makes theology vulnerable to empirical attack. Non-realists may go on (indeed, have gone on) to suggest that realism is not even religiously serious or significant: it makes religion a search for some object in the world, instead of focusing on our response to the world. Theological discourse and associated religious practice is to be valued rather for its ability to shape our attitudes to others and to the world, and to reach an awareness of our responsibilities. Realists, for their part, may complain that non-realists give us nothing to hope for, and make religious language systematically misleading. They may also point out that the moral and spiritual benefits that religion brings cannot be detached from commitment to a world view. It is precisely through attempting to align our will with that of a real God that we come to recognize our responsibilities, and come to a right attitude towards the world. Arguments on both sides seem to have considerable force, and so we find ourselves facing an antinomy. What, then, would our position be? It would, effectively, be that of semantic agnosticism with respect to theological discourse. And if we are willing to accept this much, that whichever semantics has the stronger claim, the use of theological language in shaping a religious life is rational and justified, and has a genuinely transformative power, then we have a distinctively religious semantic agnosticism, that is, an agnosticism that can be combined with a religious life. So is semantic agnosticism declining to give any kind of semantics to theological sentences? It would be better, I think, to conceive of it as offering a disjunctive semantics. For either the realist truth-conditions obtain or they do not. If they do, then on a realist semantics, the appropriate theological statements come out true. Under those conditions, the realist semantics would be the one we would wish to adopt. But if the realist truth-conditions do not obtain, then, to

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

40

Agnosticism

preserve the religious value of the discourse, we should think of it as having fictionalist truth-conditions. On that semantics, the appropriate statements also come out true, that is, true-in-the-fiction. The trouble is that we do not know whether the realist truth-conditions obtain or not. But we can, it seems (having ruled out the non-realist rivals to fictionalism), commit to a disjunctive semantics, as follows: ‘God is F’ is true if and only if: EITHER (i) the realist truth-conditions obtain, OR (ii) the realist truth-conditions do not obtain, but according to the theological fiction, God is F. Although we do not know which of (i) or (ii) obtain, we know that one of the disjuncts does, and so ‘God is F’ comes out true. We thus arrive at religious semantic agnosticism, but neither truth agnosticism nor existential agnosticism over theological statements. We are not, that is, agnostic over the truth-value of ‘God is F’, for the reason just given. And, as a result, we are not agnostic over the existence of God, since the truth of ‘God exists’ follows from ‘God is F’. This, to some extent, mirrors the situation I have suggested we are in vis-à-vis modal truths (we are confident that we can assert them, but do not know which truth-conditions obtain), and which we might be in with respect to moral truths. We can, at least, define a semantic agnosticism with respect to theological statements. This is not to say, of course, that is actually a viable position. Let us now consider a number of possible objections.

IV. A CONFUSED AMALGAM? DEFENDING RELIGIOUS SEMANTIC AGNOSTICISM I envisage three kinds of attack on the position we have just articulated: (1) It is really just a form of truth (and existential) agnosticism, dressed up in different language. (2) It is incoherent, since whether something is fictional or not requires fictive intent, and we cannot be agnostic about our own intentions.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

41

(3) That, though it is both coherent and distinct from standard agnosticism, it has the advantages neither of realism nor of fictionalism. Let’s begin with the first of these (where else?). To expand it a little: the suggestion that we are retaining the truth of ‘God is F’ whether or not the realist truth-conditions obtain is disingenuous. For there is all the difference in the world between being really true, and being fictionally true. If the realist truth-conditions do not obtain, then ‘God is F’ is really false, even though it may be fictionally true. So ‘semantic agnostics’ will still be agnostic about whether or not ‘God is F’ is really true. So why not say that this is just good old-fashioned truth (and existential) agnosticism, with the additional concession that were we to treat religious discourse and activity as a game of make-believe, it would still have a point? In reply, it can be conceded at once that there is a significant difference between being true of reality and being true of (or in) the fiction. But this is not a difference between two kinds of truth; rather, it is a difference in what is being described—real or fictional individuals. ‘Jane Eyre is ill-treated by Mrs Reed’ is true in just as robust a sense of ‘true’ as ‘Children were exploited in Victorian factories’. But whereas the second concerns really existing individuals, the first concerns fictional characters. If we supposed that this indicated two different kinds of truth, then we would have to say that there is a difference between modal statements being ‘abstractly true’ and being ‘concretely true’, rather than framing the difference in terms of what those statements are true of: abstract possibilia, or concrete possibilia. So statements of the form ‘God is F’ come out true on semantic agnosticism, in a univocal sense of ‘true’, but what those statements are true of—a real God or a fictional one—is unknown. Acknowledging this, the critic may then say that nothing is really added by including a fictionalist semantics as one of the disjuncts. Why not say that, if the realist truth-conditions do not obtain, then statements like ‘God exists’ turn out to be false, but that even false statements can play a role in shaping our lives in beneficial ways? Perhaps, for example, it is false that we are free in some libertarian sense, but it might still be a useful (indeed indispensable) falsehood. There is, however, genuine benefit in allowing for a fictionalist semantics to come into play, should the realist truth-conditions fail to be fulfilled. For fictions have their own logic. From the fact that

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

42

Agnosticism

p entails q, it does not follow that ‘fictionally p’ entails ‘fictionally q’: that might depend on the fiction. For example, let us suppose that Holmes and Watson travel from London to Edinburgh in one of their fictional adventures. Actually travelling from one to the other entails traversing a continuous spatial line between them. Does it follow that there is such a series in the fiction? No: this need play no role in the story. There need be no answer to the question: ‘Where, in the fiction, were Holmes and Watson exactly one hour before their arrival in Edinburgh?’ Of course, obvious inferential connections will still hold (Holmes and Watson are human beings, and so have the usual physiological constitution of human beings), but less obvious ones may not. So what might generate paradox outside a fiction will not necessarily do so within the fiction. That is of some significance when it comes to theistic matters. Consider, for example, the idea of a triune God, or an incarnate God: it has been argued that both of these entail contradiction. Supposing that to be so (which of course is disputed), it does not follow that those same ideas in the context of a fiction would generate contradictions. Since, then, semantic agnostics include a fictionalist disjunct in their semantics of theological statements, they will be more cautious than realists about the inferences they draw from those statements. Turning now to the second objection: if fiction requires fictive intent, how can we be agnostic over whether or not the discourse we are engaging with has a fictionalist semantics? In a famous (and possibly exaggerated) episode of media history, Columbia radio broadcast on 30th October 1938 an adaptation of H.G. Wells’ War of Worlds, narrated by Orson Welles. It began with what was presented as a straight news broadcast of alien invasion, allegedly causing panic amongst listeners, who thought that it was a genuine news bulletin, and Earth really was being invaded. When they discovered their mistake, they might well have thought ‘It was just make-believe’, but surely they would not have thought ‘I was just make-believing’. They knew perfectly well that they were not. But if we always knew whether we are make-believing or not, then it surely makes no sense to say that we do not know whether our religious discourse has a fictionalist semantics or not, unless all that it mean is that we have not decided whether to employ a fictionalist semantics. Until we do so decide, of course, our utterances within that discourse can have no assignable significance. The reply to this objection is that the semantic agnostic has decided on a semantics, but it is a disjunctive semantics. So there is fictive

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

43

intent, insofar as there is an intention for the fictional semantics to operate should the realist truth-conditions fail to obtain. The problem remains, however, of what attitude the agnostic should adopt whilst actually engaging in religious discourse and practice, for treating something as real is very different from playing a game of makebelieve. Although a disjunctive semantics can be given for the discourse, how does this guide the agnostic’s actual engagement with it? It seems that full engagement is ruled out. But this need not be so. In terms of attitude, the appropriate one may be to treat the engagement as one of make-believe, with the proviso that some elements of the game may be real. Some games of make-believe, for example, involve genuine, and not merely virtual, danger. So what of the third objection, that semantic agnosticism has the advantages neither of realism nor of fictionalism, but the disadvantages of both? I would like to suggest that it is the other way around. Standard (truth and existential) agnosticism adopts a realist semantics, and in so doing, gives a significant hostage to fortune, as it makes the religious significance of the discourse (and associated practice) depend on the realist truth-conditions actually obtaining. The standard agnostic can still engage in religious thought, talk, and activity, but it is likely always to be tentative, experimental.5 The hope is that the language and activity will be vindicated by the fact that they are, respectively, true and (real) object-directed. But if they are not, then this particular shot in the dark will have failed to hit its target. The semantic agnostic, by contrast, does not necessarily tie the religious significance of what is said and done to the truth (as the realist conceives of it) of what is said, and the real object-directedness of what is done. In the absence of a transcendent deity, the game of make-believe still has religious significance, in the senses identified above, and can be engaged in for precisely those purposes. The semantic agnostic need not avoid saying the words of the creeds: the propositions of professed belief are at least true in the fiction. Nor need the agnostic avoid acts of worship: they have at least a fictional object. Does semantic agnosticism have any advantages over fictionalism? The fictionalist need not, like the realist, fear falsification: the game of make-believe is immune from actual fact. So the fictionalist can 5 Agnostic attitudes to religious thought and observance, and the realist assumption of agnosticism, are explored in Le Poidevin (2010).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

44

Agnosticism

engage in religious language and activity in the confidence that it is true and object-directed. But, although it is the case that the discourse within the game is not made false by the existence of a transcendent God (for, after all, that discourse does not say anything about how things actually are), it is, arguably, made guilty of missing the truth. This is what William James professed to fear about adopting the agnosticism of Huxley and Clifford: that by not committing to belief, he would be in danger of missing the most tremendous truth imaginable.6 A pure fictionalism, then, is more likely to appeal to someone who thought that realism was not even viable, that it led to paradox, or put theology in conflict with contemporary science. But given any uncertainty about religious realism, semantic agnosticism will seem a more attractive position, precisely because it remains sensitive to the possibility of the kind of transcendental reality described by the realist. In summary, then, semantic agnosticism aims to retain the advantages of both realism and fictionalism, while avoiding the disadvantages of both. It avoids realism’s risk of missing the truth, but it does not (unlike realism) make the religious significance depend solely on a transcendent reality.

V. RETURN TO THE BOTTOM OF THE GARDEN So let us go back to the summer of 1968 and try to make sense of what two boys thought they were doing to the summerhouse. Was the ramshackle old building just a prop in an extended game of makebelieve? Or the beginnings of a projected (though clearly doomed) trip into space? Or something else? I will court derision, perhaps, if I suggest that we were, in effect, semantic agnostics about the project, because that suggests a degree of theoretical reflection we were not capable of. But what the term ‘agnostic’ might signal here is an absence of such reflection. We didn’t (I suspect) want to take a stance on exactly what it was we were doing, neither treating it wholly as pretence, not yet committing ourselves to the alternative. It could 6 The source, of course, is James’ famous essay, ‘The Will to Believe’ (James 1905 [1897]). For the side he was opposing, see Huxley (1904 [1889]) and Clifford (2001 [1877]).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

The New Agnosticism

45

have been either—did we need to think about it? Perhaps we wanted to enjoy the advantages both of make-believe (making it the case that this bit of apparatus was the navigation system true by fiat) and of realism. We did not want to confront the dangers attendant on space exploration (particularly at this early date in the development of manned rocketry). But we wanted—and indeed achieved—the excitement that a project with a genuine, and not merely feigned, purpose provided. So we hovered between make-believe and realism, and certain kinds of confrontation from critics might have pushed us one way or the other. There is, I propose, at least an analogy between this and religious semantic agnosticism. The difference is that the latter is a position adopted after critical reflection, and an explicit commitment to a disjunctive semantics. But having taken this critical step we could, as it were, forget it, or bracket it, and allow our thought and activity to hover between make-believe and realism. It is sometimes objected that theories of religious discourse, particularly of a non-realist nature, are over-sophisticated, and would actively disrupt religious engagement. But what I am envisaging here is a rather more indeterminate state of mind, where these kind of reflections are suspended. A state of mind not unlike—indeed, strikingly like—that of children absorbed in their fantasies. And that a childlike state of mind is best suited to a religious life is not an unfamiliar theme: ‘Whosoever shall not receive the kingdom of God as a little child shall in no wise enter therein’ (Luke 18:17). If religious semantic agnosticism has any merits, as I have suggested here, then it deserves a simpler and more memorable name. I suggest ‘The New Agnosticism’. The old summerhouse, by the way, is still at the bottom of the garden. But it belongs to another family now. I hope it fires the imagination of the children that play there now as much as it did ours then. BI B LI OGR APHY Braithwaite, Richard. ‘An Empiricist’s View of the Nature of Religious Belief ’. In Philosophy of Religion, edited by Basil Mitchell (Oxford: Oxford University Press, 1971 [1955]), pp. 72–91. Bultmann, Rudolf. ‘A Reply to the Thesis of J. Schniewind’. In Kerygma and Myth: A Theological Debate, edited by Hans-Werner Bartsch, tr. Reginald H. Fuller (London: SPCK, 1972).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

46

Agnosticism

Clifford, W.K. ‘The Ethics of Belief ’, Contemporary Review 29, pp. 289–309. Reprinted in Philosophy of Religion: An Introduction with Readings, edited by Stuart Brown (London: Routledge, 2001 [1877]), pp. 163–8. Cupitt, Don. Taking Leave of God (London: SCM Press, 1980). Divers, John. ‘Possible-worlds Semantics without Possible Worlds: The Agnostic Approach’. Mind 115, no. 458 (2006), pp. 187–226. Eshleman, Andrew. ‘Can an Atheist Believe in God?’ Religious Studies 41, no.2 (2005), pp. 183–99. Huxley, T.H. ‘Agnosticism and Christianity’. In Collected Essays, Vol. V: Science and Christian Tradition (London: Macmillan and Co., 1904 [1889]). James, William. ‘The Will to Believe’. In The Will to Believe and Other Essays in Popular Philosophy (New York: Longmans Green and Co., 1905 [1897]). Le Poidevin, Robin. Agnosticism: A Very Short Introduction (Oxford: Oxford University Press, 2010). Le Poidevin, Robin. ‘Playing the God Game: The Perils of Religious Fictionalism’. In Alternative Conceptions of God, edited by Yujin Nagasawa and Andrei Buckareff (Oxford: Oxford University Press, 2015), pp. 179–91. Mackie, J.L. Ethics: Inventing Right and Wrong (Harmondsworth: Penguin, 1977). Phillips, D.Z. Religion without Explanation (Oxford: Basil Blackwell, 1976). Wittgenstein, Ludwig. Lectures and Conversations on Aesthetics, Psychology and Religious Belief, edited by Cyril Barrett (Oxford: Blackwell, 1966).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

2 What is Wrong with Agnostic Belief? Yuval Avnur

What makes an agnostic neither a theist nor an atheist? This is sometimes taken to be a matter of belief about some relevant religious hypothesis; call it ‘h’. A theist believes that h is true, while an atheist believes that h is false. On this way of carving up the territory, an agnostic is a non-believer; an agnostic lacks belief that h and lacks belief that not-h. Non-belief is in turn sometimes understood as a ‘suspense of judgement’, though that notion has proved difficult to analyse.1 There is another way to understand agnosticism. From a philosophical, or at least an epistemological perspective, we are interested in evidence and arguments for beliefs, not just beliefs themselves. Though theism is a metaphysical position, that h (e.g. that there is a God), a theist thinker presumably also thinks that the evidence (or arguments, which I count as a sort of evidence) ultimately indicates that h is true. And an atheist thinker presumably thinks that evidence ultimately indicates that h is false. Accordingly, an agnostic thinker should hold that evidence doesn’t ultimately or sufficiently indicate either way about h. There are various ways for evidence to fail to ultimately indicate something, and below I will focus on a particular one. But even prior to that disambiguation, the distinction between

1 See Monton (1998) and Friedman (2013) for critical discussion of some attempts to formalize an agnostic attitude. I briefly discuss Friedman’s positive proposal at the end of this chapter.

Yuval Avnur, What is Wrong with Agnostic Belief ? In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0003

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

48

Agnosticism

agnosticism and the first two positions with respect to evidence about h should be clear. So far I have left out a sort of fideism, or the attitude of believing that h while also believing that the evidence fails to ultimately indicate whether h is true. My main question concerns this sort of position. A couple of preliminary remarks about agnosticism as a position about evidence will be helpful, since agnosticism is so often assumed to be simply a matter of withholding belief. First, construing agnosticism as having this epistemological component is certainly consistent with the way in which the term was originally introduced: T.H. Huxley2 introduced the term in the nineteenth century by describing agnostics as ones who ‘like [himself], confess themselves to be hopelessly ignorant concerning a variety of matters, about which metaphysicians and theologians, both orthodox and heterodox, dogmatise with the utmost confidence’. Regarding oneself as ‘hopelessly ignorant’ with respect to h involves, at least, thinking that the evidence does not (perhaps could not, since the situation is ‘hopeless’) settle whether h. Second, lack of belief about h can be due to things like an abundance of caution, stubbornness, failure to even consider evidence, or one’s regarding h as so fantastic, or so desirable, that one hesitates to trust what seems to one to be even a fairly solid case in favour of h. Surely, these ways of lacking belief about h are not the only ways to be agnostic. There is also a more principled, philosophical sort of agnosticism. Taking a philosophically agnostic position is not merely a matter of lacking belief for any reason; it involves lacking belief about h because, roughly, one thinks that it is unknowable whether h.3 When this epistemological, sceptical position is fleshed out in more detail (in section I), its worthiness of the title ‘agnosticism’ should become clearer. I have specified two sorts of agnosticism: non-belief (or suspense of judgement) about h, and belief about the insufficiency of our evidence about h. I do not aim to adjudicate which of the two (if either) should count as the most correct way of understanding agnosticism. Instead,

2

Huxley (1884). For a different approach, see Schellenberg (2009) and (2013), pp. 100–106, where a notion of ‘imaginative faith’ is developed in the place of what I am calling ‘agnosticism’. 3

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

49

my aim is to consider a key question about the relation between the two. Does the position that the evidence supports neither h nor not-h commit one to non-belief about h? It may well seem perverse to go to the trouble to develop a negative view of a belief ’s epistemic credentials, and then to hold the belief anyway—and this seems equally perverse whether we have in mind belief in theism or belief in atheism. But what, exactly, is wrong with believing something that you are agnostic about in the second, epistemological sense? What is wrong with believing h (or not-h) while also believing that there can be no evidence about h? That will be my main question, and I will call belief in a proposition that one is also epistemologically agnostic about ‘agnostic belief ’. I begin in section I by clarifying the sort of agnosticism I have in mind, namely the second, epistemological version. In sections II and III, I explain and, to some extent, defend two assumptions that facilitate answering the main question, ‘What is wrong with agnostic belief?’. I then consider what I take to be the three main answers: such a belief would be unjustified for lack of evidence (section IV); one’s set of attitudes would in this case be problematically akratic or incoherent (section V); and one would be committed to a judgement—a socalled ‘Moore-paradoxical’ judgement—the absurdity of which reveals the irrationality of being an agnostic believer (section VI). After arguing against each of those answers, I consider, in section VII, some other features of defensible agnostic belief. Before proceeding, a clarification about the main question is in order. There are different senses in which a state or attitude can be wrong. I will consider what is ‘wrong’ with agnostic belief exclusively from an epistemic perspective. This means that when we attempt to explain what is wrong with agnostic belief, we are attempting to understand a deficiency in the belief ’s, or in the believer’s, accuracy with respect to the truth, from the subject’s perspective. The relevant aspect of the subject’s perspective is to be understood in terms of her evidence. So the question, clarified, is: What is wrong with agnostic belief, from an epistemic perspective, which is one that is concerned with accuracy relative to the evidence? In pursuing the question in this way, I will assume that a belief ’s or a believer’s ‘rationality’ is a measure of that belief ’s or that believer’s accuracy relative to the believer’s evidence. Accordingly, the answers to the question that I will consider all take the following form: What is wrong with agnostic belief is that it is always (epistemically) irrational.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

50

Agnosticism I. SECOND-ORDER AGNOSTICISM

Recall that an agnostic stance on some hypothesis, h, is supposed to be contrary to both the stance that h is true and the stance that h is false.4 But the sort of agnosticism I will develop here is not defined as a failure to take a position on h’s truth. Rather it is a position with respect to taking a position on h’s truth. This second-order attitude is explained in general terms in this section. Though there are differing candidate attitudes that might constitute some other, neutral stance deserving of the name ‘agnosticism’, here I focus on one, secondorder way of being an agnostic. I will briefly mention some others, at the end of this section, for contrast. Suppose that you believe that you have absolutely no evidence that bears on a hypothesis, so that you do not even know what probability to assign it. When you have this belief, you are not merely ignorant, since you believe that you are ignorant in the relevant respect (we surely are unwittingly ignorant about many things that we have never even considered). This belief does not yet constitute the agnosticism I want to focus on, but it is a start. Agnosticism has something to do with believing that you, perhaps that we all, are ignorant about something. To get from believing in your ignorance about something to the sort of agnosticism I want to discuss requires specifying the sort of ignorance one believes in. Roughly, the agnosticism I have in mind consists in the belief that no foreseeable source of evidence could bear on h sufficiently to justify belief about h. However, some further clarifications are needed, and this agnosticism needs to be further narrowed down. First, you must believe that your total evidence is insufficient. This means you lack background (inductive or statistical) evidence about the hypothesis. For example, to borrow Russell’s famous orbiting teapot example,5 you may lack direct evidence that there is a teapot orbiting the Earth, and yet possess some background evidence about what teapots are, where they come from, what it takes to launch something into orbit, whether astronauts take teapots with them and dispose of them while in orbit, and so on. This was not the 4

Rosenkranz (2007) focuses on agnosticism as a stance concerning our knowledge, whereas I will primarily be discussing justification. Otherwise, the sort of agnosticism I will focus on here is similar to the sort he discusses. 5 Russell (1952).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

51

point Russell was trying to make with this example. In fact, he seemed not to recognize how our background evidence bears on the hypothesis. But it can be used to illustrate the importance of distinguishing total from direct evidence. The orbiting teapot would presumably not be a case in which you are an agnostic in the present sense, since you think that your evidence—though perhaps not direct or observational evidence—bears significantly on the hypothesis. We think we have sufficient evidence to show that it is extremely unlikely that there is such a teapot; we are atheists about orbiting teapots.6 Another way for background evidence to bear on a hypothesis, when one lacks direct evidence about it, is for one’s understanding of the case to partition the possibilities in such a way as to determine a probability. For example, consider the hypothesis that the next time you roll a fair die ten times, it will land ‘6’ each of those ten times. You understand what it is for a die to be fair, and so you ‘see’ that this hypothesis is very unlikely to be true. We may wish to count this as evidence against the hypothesis. Regardless of whether we count it as evidence against the hypothesis, this is a case in which, though we lack direct evidence about the hypothesis (since the rolling of the die has not yet occurred and is stipulated to be fair), we should presumably not be agnostics about it, at least not in the sense that I will be discussing. Instead, we should be fairly confident that it is false, and we should believe that this confidence is justified. Again, we lack direct evidence, but we are atheists.7 So far, I have specified that an agnostic believes that her total evidence—including her understanding of the case—fails to justify belief, or some other significant level of confidence about the hypothesis. But furthermore, an agnostic also believes that none of her current or foreseeable sources of evidence could produce such evidence about the hypothesis. For example, I currently have no evidence about whether space suits of the sort that astronauts wear at the

6

Similarly, if you are agnostic, then you think that the fact that you lack direct evidence about the hypothesis does not itself confirm or disconfirm the hypothesis, since you would not expect to have any evidence about the hypothesis whatever its truth-value is. I lack space to discuss such cases here. See Sober (2009a) and Avnur (forthcoming). 7 I have avoided bringing up the principle of indifference in this discussion. That principle will not be relevant to the cases I end up discussing below, and so will not help to determine a probability for the hypotheses about which I suggest we should be agnostic.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

52

Agnosticism

space station are comfortable when worn in orbit. I’ve never read about this, never interviewed an astronaut, and I’ve never been in such a suit at the station myself. But I know that I could, in principle, find out: maybe there is or will soon be some existing literature on it, maybe I could interview a current or past astronaut, and maybe I could get hold of a suit and join some future expedition to the space station. This is not something that is likely to happen. But it is something that my sources of evidence could, in a foreseeable (though unlikely) way bear on. So I should not be agnostic about this, at least not in the sense that I will be discussing. This seems an acceptable interpretation of agnosticism. For, presumably, a philosophical agnostic about a religious hypothesis does not typically think that if we just tried harder, or investigated further, or called the right person, we would find evidence about it. At any rate, at least one recognizable sort of agnostic thinks that, in principle, we do not have access to good evidence about h, and this is the position I have in mind. In section III, below, I suggest that we are sometimes justified in such a stance with respect to some religious hypotheses. As I noted above, the belief that constitutes this kind of agnosticism has a second-order character. It is about what your evidence cannot justify, or about your epistemic situation with respect to the relevant hypothesis. Thus, I call it ‘second-order agnosticism’. Specifically, second-order agnosticism about some hypothesis is the belief that no (first-order) doxastic attitude about the hypothesis can be justified by the evidence (as described by the specifications above). I understand doxastic attitudes to include at least belief and disbelief (i.e. belief not-), as well as degrees of confidence about the relevant hypothesis. The ‘can’ is meant to signify that this applies to what your current and foreseeable sources of evidence could supply (it is not meant to imply that it is a necessary fact that you will always lack evidence from any possible source). I will discuss one important implication of the possibility of there, someday, somehow, being evidence in the last section, section VII. But for the most part it will not matter for my main question, which concerns agnostic belief. Finally, note that second-order agnosticism is a belief, and as such it could turn out to be false. Second-order agnosticism is worth distinguishing from other attitudes that might be considered agnostic, but which are more direct or first-order. Clearly it is distinct from any first-order attitude since it is a second-order attitude, but, further, first-order agnosticism does not

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

53

necessarily commit one to second-order agnosticism. There are hypotheses and evidential situations such that some first-order type of agnosticism may be appropriate while second-order agnosticism is not (and vice versa). Here are a few instances of this. First, simply having no attitude about h might be interpreted as agnosticism about h. But we have this (non-)attitude about hypotheses which we have never even considered, and clearly we do not and should not believe second-order agnosticism about hypotheses we have never even considered. Second, having a .5, or middling, degree of confidence in a hypothesis might be thought to be a sort of agnosticism about it. But one might have this attitude on the basis of evidence that is equally weighted for and against the hypothesis. If so, then second-order agnosticism would not be appropriate, since evidence could bear on the hypothesis. That is, one’s sources of evidence are such that they could produce evidence sufficient for justified belief (say, in the absence of the other, competing evidence that one actually happens to have). Third, one might think that considering, and then withholding, belief about a hypothesis constitutes a type of agnosticism. But one could be in this state without properly being a second-order agnostic. For example, one might withhold belief while the investigation is ongoing, while one is considering the weight of various items of evidence, or watching the evidence come in. In such a case, of course one may well believe that one’s sources (or foreseeable sources) could sufficiently bear on the hypothesis, so one is not a second-order agnostic. So far, I have suggested that the having of various first-order ‘agnostic’ attitudes towards a hypothesis does not commit one to second-order agnosticism. My main question now concerns being a second-order agnostic while not withholding belief in the relevant hypothesis. That is, is it necessarily irrational to be a second-order agnostic about h while having a belief or confidence one way or another about h? I am calling this combination of attitudes—secondorder agnosticism about a hypothesis and belief or confidence about that hypothesis—agnostic belief. Before considering the question, though, two assumptions must be made, explained, and defended. In section II, I consider whether agnostic belief is even possible. In section III, I consider whether second-order agnosticism itself can ever be justified, with respect to some religious

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

54

Agnosticism

hypotheses. After that, in sections IV VI, I will consider whether agnostic belief is always irrational.

II. FIRST ASSUMPTION: AGNOSTIC BELIEF IS POSSIBLE In considering the rationality of agnostic belief, I will assume that agnostic belief is logically, conceptually, and psychologically possible. I think that these assumptions are plausible but I will not have space to argue thoroughly for that. Instead, I will just make the assumptions explicit and gesture towards their motivation. After this, I will also suggest that these assumptions may not really be necessary, though they are useful. I assume that agnostic belief is logically possible. This seems a safe assumption because there is no formal contradiction in attributing to some subject both second-order agnosticism and belief: the belief that constitutes second-order agnosticism about h is distinct from belief that h, by definition. Furthermore, the truth of second-order agnosticism about h is compatible with both the truth and the falsity of h. So there seems to be no reason to hold that it is logically impossible for someone to hold an agnostic belief. I assume that agnostic belief is conceptually possible in the sense that the concept of belief does not rule out agnostic belief. Some hold that, in order to be a belief, an attitude must be susceptible to change due to new evidence.8 But this account of belief does not make agnostic belief impossible. For, it does not follow, from one’s being a second-order agnostic about h, that one’s attitude towards h is not susceptible to evidence, even by one’s own lights. For instance, if some unforeseeable source of evidence were to arise, the agnostic believer’s belief that h might change, and she would no longer be a second-order agnostic about h. For example, if an agnostic believer learns that her second-order agnosticism is mistaken, and, say, there is actually evidence against the hypothesis, she could (presumably would) drop her belief that h. And, to take another example, even if 8 For example, Gendler (2008) contrasts belief in this way with what she calls ‘alief ’, and Van Leeuwen (2014) contrasts belief in this way with what he calls ‘religious credence’.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

55

one’s second-order agnostic belief about h is true, and no current or foreseeable source of evidence can bear on h, it may still be true that, supposing that one were to get evidence (as is impossible in this case), one would change one’s attitude about h. The disposition to be sensitive to evidence might be there even in the absence of any possible evidence. The agnostic belief is, in that case, much like a fragile glass in a world filled with pillows.9 Granting that agnostic belief is logically and conceptually possible, there is still the question whether it is (humanly) psychologically possible. That it is psychologically possible is really the main assumption, since this possibility entails that agnostic belief is both logically and conceptually possible. It is not only the strongest assumption, but also the most complex and contentious. I will discuss this assumption in more detail, below, when discussing akrasia. For now, I will simply assume that agnostic belief is psychologically possible and point out a few things about it. First, notice that second-order agnosticism is about what can be justified by the evidence. So, anyone who thinks that, eventually, we must believe something without evidence because evidence must come to an end, must hold that agnostic belief is psychologically possible. Some popular forms of foundationalism seem committed to this view:10 they think that it is possible to believe something when one also believes that there is no evidence for it. Similarly, those who think it is possible (though perhaps irrational) to believe something on blind faith already think it is possible to be an agnostic believer. Less obviously, one might think that cases of self-deception, in which a person believes that evidence does not justify some belief and yet still holds that latter belief, provide a third type of psychologically real case. Second, agnostic belief has been assumed to be psychologically possible throughout much of the history of philosophy, at least by some prominent philosophers. On one interpretation, Kant’s ‘practical postulates’ are beliefs (or, at least, commitments) one could have 9 See Ribeiro (2011) for discussion, historical overview, and rejection of the view that akratic beliefs are impossible. 10 Wright (2004) and White (2006) have defended this sort of view in the context of scepticism about the external world. Ribeiro (2011) holds that agnostic belief is actual in the same context, but without committing to foundationalism. Others, such as Plantinga (2000), have rejected evidentialism, and are also committed to the psychological possibility of agnostic belief, in the context of religious belief.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

56

Agnosticism

while recognizing that reason does not support them.11 Similarly, Hume seems to have held both that some beliefs cannot be supported by any reason or evidence and that having such beliefs is naturally unavoidable. Sextus arguably also held that we must assent to things without holding that there is any (unopposed) evidence for them. So this assumption is in some good company. Finally, note that, even if the assumption is wrong, and we are not able, psychologically, to hold agnostic beliefs, we may still ask an abstract, ideal, normative question about it. That is, we could still ask whether, were it possible to have such a combination of attitudes, that would be rational. This would not be an unprecedented sort of question, and it can still be an important one in this case. It is not unprecedented because, for one example, Bayesian or ‘formal’ approaches to epistemology idealize beyond what is psychologically possible. It is held, for example, that being logically omniscient is rationally ideal, despite the fact that this is psychologically impossible.12 It is still interesting to consider the rational status of agnostic belief, even if it turns out to be psychologically impossible, because that would shed light on the status of our actual state when we are second-order agnostics about something. That is, the status of agnostic belief can help to determine what we take the status of agnostic non-belief to be. It may, for example, turn out to be epistemically no better than agnostic belief, even if agnostic belief is psychologically impossible.

III. SECOND ASSUMPTION: SECOND-ORDER AGNOSTICISM CAN BE JUSTIFIED So far, I have assumed that agnostic belief is possible. I will also assume that there are situations in which it is rational, or justified, to be a second-order agnostic in the first place. This is a helpful assumption because such circumstances, in which second-order agnosticism is justified, will presumably place some constraints on one’s first-order attitude about the hypothesis in question. Furthermore, it is worth 11 12

See Chignell (2007). See Christensen (2004) for some discussion of this.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

57

considering cases in which second-order agnosticism is justified in order to further motivate our question about agnostic belief. After all, who cares if it is rational to believe something that one is second-order agnostic about if it is never rational to be second-order agnostic about anything? Finally, note that it is unnecessary to argue, or even assume, that all or even all standard religious hypotheses are such that secondorder agnosticism is sometimes justified about them. The question is whether agnostic belief is always irrational. If it can be shown that one significant (i.e. not overly contrived) example of justified second-order agnosticism might in fact be non-wrongly held in conjunction with a belief in the hypothesis, this will show that the answer is ‘no, it is not always wrong to be an agnostic believer’. This may well shed light on agnostic belief in general. In discussions of religion, an obvious and classic hypothesis about which agnosticism seems appropriate is one according to which there is a God that is transcendent in the sense that we cannot get evidence for it during our lives in this world. The existence of this God would matter to us, presumably, at least in part because it would make some sense of our existence, and perhaps hold some promise of an afterlife for us. One can even leave some room for interaction between us, the creatures of this world, and such a God, in the form of miracles that have happened long ago or in faraway places. Such miracles, though, must not have the capacity to provide us with evidence for such a God’s existence, or else belief, rather than second-order agnosticism, will be justified by the evidence. Thus, the requisite transcendence is epistemological rather than strictly metaphysical; we should think of this sort of hypothesis as one according to which an epistemically transcendent God exists. This seems, by definition, something about which we are justified in being second-order agnostics. A similar hypothesis posits some form of divinity or other, without specifying that it is a god. We can think of this as the hypothesis that an epistemically transcendent divinity exists. By hypothesis, we can have no evidence either for or against it. So, it seems that secondorder agnosticism is justified.13 On such hypothesis is James’ religious hypothesis in The Will to Believe:

13 Do explanatory considerations of simplicity and elegance tell against this hypothesis? I think not, but I cannot defend that here. See Sober (2009b).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

58

Agnosticism

the best things are the more eternal things, the overlapping things, the things in the universe that throw the last stone, so to speak, and say the final word . . . [and] we are better off even now if we believe [this] to be true.14

James famously held that the ‘intellect’ is silent on whether we should believe this hypothesis, in the sense that none of our evidence could possibly indicate whether it is true (at least independently of a belief we may have in it). Another, similar hypothesis can be found in Schellenberg’s work, under the titles of ‘Ultimism’—‘there is a metaphysically and axiologically ultimate reality (one representing both the deepest fact about the nature of things and the greatest possible value), in relation to which an ultimate good can be attained’15—and ‘T-ism’—‘Reality is transcendent—more than or other than the arena of mundane events or . . . anything physical or natural’.16 It is hard to see how our current or foreseeable sources of evidence could bear on such hypotheses. So, presumably, second-order agnosticism about such hypotheses could be justified. One might want to argue that second-order agnosticism is also justified with respect to more traditional theistic hypotheses. And there may also be secular hypotheses about which second-order agnosticism is justified.17 In assuming that some such hypotheses are plausible targets for second-order agnosticism, I do not mean to make any contribution to a debate about whether this is so. For example, I will not make any attempt here to contribute to the theological tenability of an epistemically transcendent yet personal God.18 Rather, I mean only to give some, albeit vague, content to the idea that there are some hypotheses about which second-order agnosticism is justified. My main question concerns the status of agnostic belief, and it will not matter, in what follows, what the content of the belief, or the hypothesis in question, entails. What matters is that one is justified in being a second-order agnostic about some hypothesis.

14

15 James (1897 [1896]). Schellenberg (2005), p. 23. Schellenberg (forthcoming), p. 9. 17 Examples of secular hypotheses include hypotheses about the positions of particles so far away that no information about them is possible at this point in the universe’s evolution, and perhaps some sceptical hypotheses about which, arguably, we can never get any evidence. 18 See, for example, Schellenbeg’s most recent book on the problem of divine hiddenness (2015). 16

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

59

What if one insists that there is no hypothesis about which secondorder agnosticism is justified? I have two points to make about this insistence. First, the idea that second-order agnosticism is never justified for any hypothesis is implausible for various reasons. Here are a few. First, setting the largely rejected logical positivist perspective aside, it is hard to imagine why one would insist that it is impossible for something to be true while we lack evidence for it, in principle. Second, it seems likely that there are some such hypotheses, since accuracy per se is not strictly a constraint upon evolution; there is no reason to think that our evolution requires accurate beliefs about all aspects of reality, but only those that are relevant to our satisfying the requirements of efficient reproduction. Third, there seem to be uncontroversial examples aside from the overtly ‘religious’ ones discussed above. Consider, for example, the continuum hypothesis. Plenty of other, widely recognized examples exist in the literature.19 The second point to make is that, if it turned out that there is no hypothesis about which second-order agnosticism is justified, that would be a remarkable and interesting fact. In order to appreciate how astonishing it would be if no hypothesis were worthy of our second-order agnosticism, we should consider what the consequences would be if some hypothesis were worthy. So, if for no other reason, we should still be motivated to figure out what second-order agnosticism entails. From here on, I will continue to use ‘h’ as a stand-in for some hypothesis about which second-order agnosticism is justified. Those who think there is some such hypothesis should understand ‘h’ accordingly, and those who think there is no such hypothesis can use ‘h’ hypothetically, in order to investigate the significance of their view.

IV. FIRST ANSWER: AGNOSTIC BELIEF GOES AGAINST RATIONAL REQUIREMENT TO WITHHOLD J UDGEMENT One might think that agnostic belief is irrational because the only justified first-order attitude to have in the absence of evidence is 19

See Gleiser (2015) for a survey.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

60

Agnosticism

withholding belief, or not believing either way. In that case, agnostic belief is unjustified in the same way that any belief held in the absence of evidence is unjustified. Here it will be helpful to assume, for the sake of argument, that second-order agnosticism is not only justified but also true in this case. This will simplify the discussion by eliminating the need to address whether, if one has a false belief in secondorder agnosticism about h, one’s belief that h is still unjustified.20 So, let us consider the following claim: granted that a subject correctly believes second-order agnosticism about h, believing that h (or not-h) is unjustified, and that explains what is wrong with agnostic belief that h. I have three independent objections to this claim. The first two concern the claim that belief that h is in this case unjustified: (a) this is not always so obviously plausible; and, (b), if it is true it is insignificant. My third objection is that: (c) even if it were true and significant that believing h is unjustified, this does not adequately explain what is wrong with agnostic belief.

(a) Is Agnostic Belief Really Unjustified? How plausible is it that, given that one’s second-order agnosticism about h is correct, one’s belief that h is unjustified? Many, perhaps most, epistemologists are foundationalists. According to some versions of that view, some beliefs are justified even though they are not justified by any evidence or by any further justified belief. Those are the foundational or ‘basic’ beliefs. Of course there are many different versions of this view, and this is usually offered as a solution to some sceptical problem. But the point is that many epistemologists are already committed to the idea that at least some beliefs are justified but not by any evidence. Without the further stipulation that they are mistaken, or that h cannot be one of these basic, justified beliefs, it does not follow from second-order agnosticism about h that h is unjustified.21 The observation that, at least according to some theories, evidence is not necessary for justification leaves open the possibility that 20

See Horowitz (2014) on higher-order defeat. In Avnur (2012a), I argue that justification without evidence is a problematic notion. I have also rejected the seeming analogy between anti-sceptical beliefs and religious beliefs, even granting that they both lack evidence (2012b). 21

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

61

agnostic belief that h is justified. Since my aim is merely to argue that the claim that h is unjustified does not itself explain what is wrong with agnostic belief, we can set aside whether such other theories are true.22 As long as we lack decisive objections against all versions of foundationalism that could apply to h, we have no decisive reason to regard h as unjustified.

(b) Would It Matter That H Is Unjustified? Set aside the reservations expressed above and grant, for the sake of argument, that agnostic belief that h is unjustified. How exactly would this explain what is wrong with believing h? The answer might initially seem obvious: it is wrong because it is unjustified. However, there are different ways for a belief to lack justification. One way is for one’s total evidence to tell against h, or for it to make h relatively unlikely. Given our definition of second-order agnosticism, this is not the case with h. Rather, in the case of h, evidence does not bear at all on the likely truth of h. There is no way, on the basis of consulting my current and foreseeable sources of evidence, to give a non-arbitrary verdict on the likelihood of h. So if belief that h is unjustified, it is not unjustified because of some verdict on its probability. Rather, it is unjustified because no probability assignment can be justified by my evidence in general (at least from any foreseeable source). This is important because there is a difference between a belief ’s being unjustified and a belief ’s being in some sense forbidden by evidential considerations. If the total evidence tells against the truth of h, then believing h goes against the evidence, and can be said to be forbidden by evidential considerations. But if my sources of evidence are totally silent on whether h is true, is belief that h also forbidden? It is forbidden only given some additional principle, to the effect that it is always forbidden to believe something unless the evidence tells in favour of it. We are currently granting that belief that h is unjustified, so we are in effect granting some such principle. But, even given the principle’s truth, it should be noted that when a belief ’s status as unjustified depends solely on some a priori principle of justification, the status should be taken with a grain of salt, for two reasons.

22

For example, Plantinga (2000).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

62

Agnosticism

First, there is the familiar Jamesian point that the aim of avoiding error is not intrinsically worthier than the aim of attaining (interesting) true beliefs. If evidence is silent on h, then the question whether it is wrong to believe h comes down to which of the two Jamesian aims is more valuable. Another way to see this is that one can be committed to accuracy, and thereby follow the evidence where it leads when it leads somewhere, and still endorse the policy of believing when evidence is entirely silent. By taking on the belief that h, one does not compromise one’s accuracy at all, since there is no reason (or evidence) to think that h is less likely than not. The idea that we should be on guard against error more than we should be adventurous in our believing is a popular one among philosophers, to be sure. The Jamesian point is that this is not itself a position that is based on evidential considerations. Since I am granting that belief that h is unjustified, I must draw a slightly different lesson from this Jamesian idea: believing h may be unjustified, but this is merely because we are prejudiced against error, because we take on some a priori principle that biases us towards non-belief, instead of favouring potential true beliefs—and it is certainly not because believing that h goes against any specific evidence or the total evidence. Given this fact about the unjustifiedness of h, and given that one’s primary consideration is accuracy rather than some particular version of accuracy (such as primarily avoiding error), one should not mind so much if, in this case, one’s belief is unjustified. It merely counts as unjustified because of the way we (according to what we are presently granting) count justification, not because accuracy or evidence itself tells against believing. A second way to see the insignificance of being unjustified in believing h is to consider some alternative notions of justification. We are currently assuming that believing h is unjustified. But suppose we invent a new notion, justification2. Justification2 is just like justification, and gives all the same verdicts in all the same evidential situations as justification does, except when evidence is silent on a proposition, in the way it is when second-order agnosticism is true of it. In such cases, while we are granting that believing is unjustified, we can stipulate that believing is not unjustified2, and it is also not justified2. Instead, justification2 is entirely silent on whether to believe. We can say that believing is a-justified2, rather than unjustified2. So, the only difference between justification and justification2 is that, when evidence is silent about a proposition, justification forbids

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

63

belief while justification2 says nothing about belief (or non-belief, or disbelief). Now consider the question why we should care more about justification than justification2. Justification and Justificaiton2 tell us the same exact thing whenever second-order agnosticism is false about a proposition, so in all ordinary cases, there is no difference between the two concepts. The only difference is that, when evidence is silent on an issue, justification2 has nothing to tell us. It ‘permits’ belief only in the sense that it does not forbid it—similarly, one might say that it is morally ‘permissible’ to wiggle your toes right now, not because morality supports it, but because morality does not say anything at all about it. This is not, then, a substantive verdict on whether to believe. Justification2 is simply silent when the evidence is silent. If we are concerned solely with truth and accuracy in belief, there does not seem to be any reason to prefer justification to justification2; there is no accuracyrelevant advantage to following the rules of justification, rather than of justification2. Therefore, given that our concern is solely with truth and accuracy, it does not matter that believing h is unjustified, given that it is also a-justified2. In other words, when the two concepts diverge, there is no good, epistemic reason to go along with justification rather than justification2. One way to understand this is along Jamesian lines: the aspect or principle of justification that forbids belief in the case of h does not itself derive from the values of truth or accuracy. Rather, it derives from a preference for avoiding error over believing adventurously. It is, in that sense, arbitrary with respect to accuracy. If what I have just argued is correct, then even if we grant that agnostic belief is unjustified, this should not matter much. The lack of justification, or even the forbiddenness of belief that h according to our notion of justification, would not explain why it is really, meaningfully wrong to hold an agnostic belief. For, the only thing wrong with agnostic belief, on the current proposal, is that it happens to be unjustified, given the notion of justification that we happen to employ. But we could have employed a different notion of justification that did not have this implication, and which would have been equally faithful to our valuing truth and accuracy in belief. Objection: epistemic justification, and rationality, cannot possibly be arbitrary in this way. It is impossible for something to be arbitrarily justified or unjustified, from the perspective of accuracy, because justification is precisely the notion we use to track accuracy (from a

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

64

Agnosticism

given subject’s perspective). So, something must be wrong with the argument above. Reply: We have a choice here. If we accept this objection, we should take it as an objection also to the assumption that it is unjustified, or irrational, to believe h in this case. If so, then its being unjustified to believe h does not explain what is wrong with believing h, since the belief is not unjustified. That is, the objector must take the argument about rationality2 as a reductio of our opening assumption that belief that h is unjustified. But we need not take a stand on whether the objector is right. The argument presented here leaves us at least with this disjunction: either belief that h is not unjustified, or else it is unjustified only because we happen to appeal to justification rather than justification2, and this is an arbitrary decision. The objection eliminates the second disjunct. But, regardless of whether it succeeds, we cannot explain what is wrong with believing h by appealing to the claim that it is unjustified (for, according to the objection, it is not unjustified, and otherwise it is merely arbitrarily unjustified).

(c) What Does Agnosticism Have to Do with It? My first point, (a), was to cast some doubt on the idea that agnostic belief is always unjustified. My second point, (b), was that even if it is unjustified, this is not a significant fact about agnostic belief, given the source or basis of its being unjustified. My third point is entirely independent of the first two, and it can appeal even to those who think that (a) and (b) are entirely wrong-headed. Our sense that there is something wrong with agnostic belief that h is not merely the sense that belief that h is unjustified. There are many ways for a belief to be unjustified even in the absence of a further, higher-order belief about it. Surely, at least part of what is wrong with agnostic belief that h is the combination of believing both that there is no evidence for h and that h is true. The second-order belief surely has something to do with what is wrong here. But the claim that the belief that h is unjustified does not imply anything about there even being a second-order belief. In other words, insofar as we think that possessing the second-order belief—being an agnostic—is significant to what the agnostic believer is doing wrong, the suggestion that belief that h is unjustified does not satisfactorily explain what is wrong. The distinctive problem with agnostic belief has not yet been identified.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

65

V. SECOND ANSWER: AGNOSTIC BELIEF IS AN AKRATIC STATE The phenomenon of akrasia is familiar in the practical sphere: it is, very roughly, when you do what you think you should not do. Philosophers have realized that there is a doxastic analogue: when you believe what you think you should not believe. Is this what the agnostic believer does, and is it always irrational? The crucial question here is what, exactly, are the two conflicting states that the agnostic believer is purportedly in? One of them is belief that h. But what is the further state that clashes with this belief in a manifestly irrational or incoherent way? As we saw above, even assuming second-order agnosticism is correct about h, and that the subject believes second-order agnosticism about h, it is not obvious exactly what the subject should think about the status of h. Given the considerations offered in section IV(a) and (b), we cannot simply assume that the agnostic thinks that one should not, all things considered, believe h. For, in the first place, it is not obvious that second-order agnosticism entails that belief that h is unjustified, and, in the second place, even if it did entail that believing h is unjustified, this seems an arbitrary verdict from the perspective of accuracy. The only uncontroversial verdict about belief that h is that it is not justified on the basis of evidence, or that it is not the case that you should believe h. We cannot assume that every agnostic believer believes much more than that about the status of belief that h. In light of the fact that the most we can say on behalf of the agnostic is that believing h is not justified, the following pairs of states are the most straightforward candidates for akratic description of the agnostic believer (in first person, in order to display the potential incoherence): (1) Believing that I am not epistemically justified in believing p, and believing p (2) Not believing that I am epistemically justified in believing p, and believing p These are situations in which one believes without justification, not against it. Consider a practical analogy. Suppose that I have no justification—and, let us specify, this is because there is no reason— to wiggle my toes right now. Suppose that I wiggle my toes. Is this akratic or incoherent? Presumably not. For, it is not the case, right

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

66

Agnosticism

now, that there is some reason to refrain from wiggling my toes (or at least I do not believe there is). So the bare fact that I do something while thinking that there is no justification to do it, as in (1), or failing to think that there is justification to do it, as in (2), is not enough to make me incoherent, or irrationally akratic. In order for a higher-order state to clash with a belief in an akratic way, the higher-order state must weigh against, or negatively assess, the belief. But, since evidence is silent on h (in a way, recall, that does not support even a 50/50 assessment of h’s chances) it is impossible that such a higher-order state is justified in this case. Or, more precisely, if it is justified, it is only arbitrarily, and so not importantly, justified. A straightforwardly akratic, or internally incoherent, state is not entailed by agnostic belief, at least not without further controversy, given the discussion in section IV(a) and (b). One might suspect that too much here rests on my previous objections, in section IV, to the view that belief that h is unjustified or forbidden. My objections were admittedly not meant to be conclusive. Rather, they purported to show that it is not clear whether belief that h is forbidden, and that if it is forbidden, it is only arbitrarily so. An agnostic believer might be convinced by my arguments, and so fail to believe that her own belief is significantly forbidden. So she might not be akratic. Still, one might think that it is also not clear whether agnostic belief should entail akrasia. A more explicit illustration of the combination of attitudes at issue here, and an assessment of the combination in light of considerations of accuracy, will be helpful. Goldman has usefully distinguished ‘ex ante’ justification from ‘ex poste’ justification.23 One has ex poste justification when one’s belief is itself supported by evidence. One is ex ante justified in believing p when p is the right thing for her, the subject, to believe. Whether one has ex ante justification, then, depends on the rest of the believer’s states and beliefs. The idea that akrasia is irrational, and that therefore the agnostic believer is irrational, is a matter of irrationality ex ante, presumably. For it is in virtue of holding the combination of beliefs, not some single belief type considered in isolation, that the subject is said to be akratic. The view currently being considered is presumably that the agnostic lacks ex ante justification because the agnostic

23

Goldman (1979), p. 21.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

67

believes that her belief that h lacks evidential support, so presumably belief that h is not the right thing for her to believe. But now notice that this ex ante irrationality, at least in this case, is not epistemic, or related to accuracy. For, all the agnostic believer has done to deserve to be called irrational is add the belief that h. And, by stipulation, the agnostic can have no indication whatsoever that h is likely to be false. So the agnostic went from rational to irrational by adding a belief which does not affect her overall accuracy from her evidential perspective. So the akrasia involved here is not epistemically irrational, or irrational from the perspective of truth and accuracy. It is perhaps not an ideally coherent state to be in, but that would not explain why being an agnostic believer is epistemically wrong. The point can be summarized as follows. Consider a system of beliefs that is coherent and includes the belief in second-order agnosticism about h. Adding the belief that h is true (or the belief that h is false) does not make this system incoherent. If it degrades the system, it does not do so epistemically, since relative to, or internal to, the system there is no believed indication that believing h decreases accuracy. There is no epistemic inconstancy in such a system since belief that h need not be forbidden according to the system, even if it is not supported or justified by the system. One final note about knowledge will be helpful here. Some philosophers24 have suggested that believing something commits one to having knowledge of it, or that knowledge is the norm of belief. The second-order agnostic holds that p is unknowable, so, according to the knowledge norm of belief, agnostic belief is irrational.25 Interpreted as a view about epistemic rationality of belief, the knowledge norm seems implausible, as others have argued.26 Though I lack space to discuss the debate at length here, one can appreciate the initial difficulty for the knowledge norm of belief by considering simple, justified, false beliefs. False justified beliefs are not knowledge, but they are still justified beliefs. For example, I could have very good evidence that my wife is a Catholic, but it could turn out that she is not. Perhaps she has been carefully hiding her true religion from me for years. If that happens, I was still justified, by all that evidence, in 24

For example, Williamson (2000); Owens (2000); and Huemer (2007). Thanks to Paul Hovda for bringing the knowledge norm to my attention in this connection. 26 For example, Littlejohn (2010) and Hawthorne, Rothchild, and Spectre (2016). 25

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

68

Agnosticism

believing she was Catholic. Given all the evidence I had, it seems it would have been irrational, from an epistemic point of view, not to believe that she is Catholic! So, although I did not know she is Catholic, it was rational for me to believe that she is Catholic. The knowledge norm seems, prima facie, false. Note also that similar views about assertion,27 though more plausible, are irrelevant to belief, in the absence of further argument. I discuss the relation between asserting and believing in section VI, in connection with Moore’s paradox.

VI. THIRD ANSWER: AGNOSTIC BELIEF COMMITS ONE TO A MOORE-PARADOXICAL JUDGEMENT Agnostic belief may still seem to involve some sort of inconsistency, because it entails a recognition that evidence fails to support what one believes. In section IV((a) and (b)) I suggested that this absence of evidential support for one’s belief—that is, the second-order agnosticism about h—does not necessarily make believing significantly unjustified. And, in section V, I argued that having both agnostic and believing attitudes is not necessarily epistemically irrational. But if one has both attitudes, and one is conscious of this, then one is in a position to judge that one is in both attitudes. This judgement can be performed to oneself or to others, in the form of an assertion, affirming that one does not endorse (or think there is evidence for) a belief that one has. The idea considered in this section is that this self-conscious judgement leads, in two steps, to an explanation of what is wrong with agnostic belief. The first step is to claim that this judgement, to the effect that one holds both agnosticism and belief about the same hypothesis, is odd or, worse, absurd. The second step is to claim that the absurdity of judging that you are an agnostic believer shows that there is something wrong with being an agnostic believer. However, both steps are independently problematic, and both would need to be correct in order for this approach to succeed.

27

For example, Unger (1984); Williamson (2000); and Lackey (2007).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

69

I will frame this discussion in terms of so-called ‘Moore-paradoxical sentences’.

(a) Is a Moore-paradoxical Sentence Produced by Agnostic Belief Absurd? Moore-paradoxical sentences are conjunctions that clearly could be true and yet it is absurd, perhaps incoherent, for someone to assert or judge them to be true.28 So perhaps one way to bring out what seems irrational about agnostic belief is to put the two elements, the agnosticism and the belief, together into a conjunction of the sort that gives rise to Moore’s paradox. In what follows I use the term ‘judge’ to mean ‘either assert to others or think to oneself ’. To take the standard example, it could be true about someone, S, that It is raining and S does not believe that it is raining But it is absurd for S to judge It is raining and I do not believe that it is raining. The puzzle is to explain why the second conjunction is absurd, given that, as we just saw, it could be true. I will not summarize the vast literature on this here. But it is possible to determine whether this will settle our question by considering some initial rough ideas about the paradox. It is tempting to say that the first-person judgement above seems absurd because, in the first conjunct, S seems to express or imply (among other things) that S believes that it is raining; and the second conjunct specifies that S does not believe that it is raining. That is one solution to this version of Moore’s paradox, if it correctly explains what is wrong with judging the conjunction. In order to gain insight into agnostic belief, we need to consider other, epistemic variations of the paradox, in which an assertion is conjoined with an epistemically relevant claim.29 For example, It is raining and S does not know that it is raining might well be true, but It is raining and I do not know that it is raining 28 29

See Coliva (2015) for a recent discussion of Moore’s paradox. See Huemer (2007) for discussion of epistemic versions of Moore’s paradox.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

70

Agnosticism

seems very odd. One natural explanation for this is that, if you do not know that it is raining, you should not assert that it is raining in the first conjunct. But, is the agnostic believer committed to any such, absurd judgement? I think not. The agnostic believer’s agnosticism implies that she does not know whether h is true. So she presumably is willing to judge that she does not know that h. This matches the second conjunct above (‘I do not know that it is raining’). But what about the first conjunct? The agnostic believer is not necessarily an agnostic asserter, to others or to herself in thought. That is, it is consistent with being an agnostic believer that one also does not, and should not, make an unqualified judgement that h. What is the relation between believing h and being willing to judge, without qualification, that h? It is often noted that, when one asserts something, one is expressing belief in that thing. That is, according to one standard theory of language, one of the main functions of assertion is to communicate beliefs. But it does not follow that one should be willing to assert anything that one believes. And, clearly, asserting ‘h’ does not merely or only express a belief that h. It does other things too. For example, according to the knowledge norm of assertion, it expresses a commitment to knowing that p.30 That is a good explanation of why ‘p and I do not know that p’ sounds so odd. It could be this other stuff that an assertion that h expresses, not the mere belief in h, that clashes with the second conjunct, ‘I do not know that h’. If so, then the mere belief in h does not clash with the judgement that one does not know that h. To appreciate this point, notice that there is more than one way to communicate, or to judge, that one believes something besides asserting it. One can say, or say to oneself, that one believes it.31 One can say ‘I believe that p’ with the aim of communicating that one believes that p. (That is not to say that this is the only thing one can communicate with ‘I believe p’.) And notice that I believe that p but I do not know that p

30

See Unger (1984) and Williamson (2000). See Moran (2001), pp. 100–7, on this point; and Chislenko (2014) for some discussion. 31

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

71

does not sound so bad. In fact, however that conjunction initially strikes you, you should not, on reflection, think it is absurd. Consider the fact that the following sounds absurd: I know that p but maybe p is false However, this assertion does not sound absurd: I believe that p but I might be wrong about that and nor does this assertion: I believe that p but maybe p is false If ‘I believe that p’ does not clash with ‘I might be wrong about that’, then ‘I believe that p’ cannot commit you to ‘I know that p’, since that does clash with ‘I might be wrong about that’. Therefore, ‘I believe that p and I do not know that p’ must not be Moore-paradoxical, at least not in virtue of the states that the two conjuncts directly describe (namely, agnostic belief). So whatever explains the oddity of P and I do not know that p it is not that believing that p commits you to knowing that p.32 Some33 hold that there is a sort of belief, ‘outright’ or ‘categorical’ belief, which commits you to practical certainty. Briefly, they think that when you categorically believe something, this ends the inquiry into it, and the possibility that it is false is off the table (absent new, further evidence that dislodges your categorical belief). You can therefore appropriately assert it, without qualification, in your own thinking and to others. So if the agnostic believer categorically believes h, then it seems she is indeed committed to the judgement ‘h and I do not know that h’, which we are granting is an absurd judgement. However, I doubt that what we usually mean, in ordinary English, by ‘believe’ is categorical belief, so I am inclined to treat this view with a huge grain of salt. Even Huemer acknowledges, for example, that ‘I believe that p, but p might be false’ does not sound problematic, or at any rate is not Moore-paradoxical.34 His explanation is that, when we

32 See Hawthorne, Rothchild, and Spectre (2016) for discussion of similar sorts of Moore sentences, also leading to the conclusion that the norm for assertion is stronger, or more demanding, than the norm for belief. 33 For example, Owens (2000) and Huemer (2007). 34 Huemer (2007), p. 153.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

72

Agnosticism

describe ourselves as believing something is this case, we have in mind a non-categorical belief; we have partial belief in mind, or a belief that consists in having a high enough degree of confidence. Even ignoring the question of what the ordinary English ‘believe’ means, we can set aside categorical belief. For our purposes, it would be interesting enough if the following state turned out to be not irrational: having a high degree of confidence in h while thinking that h is unknowable because there is no evidence for it (in the sense that second-order agnosticism is true about it). This is not a ‘retreat’ to a less interesting view than was originally advertised. For one, merely being confident in h, or thinking that h is probably true, does what we usually take beliefs to do. It can explain someone’s actions, for example: he climbed the tree because he was pretty confident that there was gold in the bird’s nest. Categorical belief is not required for explaining behaviour in the way that beliefs as we usually think about them do. Second, most other beliefs are just like the non-categorical agnostic belief with respect to our unwillingness to unqualifiedly assert their content. For example, when I am discussing with someone why I believe, or the reasons for believing, that the climate is changing, I am not going to just assert, without qualification, that it is changing, even though I do believe that the climate is changing. Similarly, I believe that I will not die completely broke. But given that I think there is some small chance that I will die broke, I will not just use the premise that I will not die broke when thinking about my children’s future. Instead, I am going to use the premise that most likely I will not die broke, or maybe that I will not die broke with the qualification that I am assuming the worst-case scenario will not happen. To sum up, I grant for the sake of argument that the epistemic Moore sentence, of the form ‘p and I do not know that p’, is an absurd judgement. But these conjuncts can be taken to express only agnostic categorical belief. This leaves open whether agnostic non-categorical belief is always irrational, and this is arguably the belief state we have had in mind all along, and at any rate deserves our attention.

(b) Would the Absurdity of Judging the Conjunction Show What is Wrong with Agnostic Belief? There is another, more general point to make here, independent of the details of Moore’s paradox. Recall that the second step was to infer,

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

73

from the apparent absurdity of judging the relevant conjunction, that there is something wrong with having the attitudes that the conjunction describes. This is a dubious inference, regardless of whether the judgement describing agnostic belief is absurd. My objection to this second step is more decisive than my objection to the first, because the objection to the first step involved consideration of some specific conjunctions and our intuitions about them. The specifics are debatable, intuitions can be pumped in different directions, and in any case perhaps new, clever conjunctions that the agnostic believer is committed to can be found. For example, consider some conjunctions that more directly describe the agnostic believer’s situation: I believe h and I lack evidence for it Probably h, and there is no reason to think that h. Perhaps to some people these sound odd. What follows from this about non-categorical agnostic belief? Not much, I will now argue, because such simple conjunctions do not specify enough about the agnostic believer’s state. Consider an analogy. I judge: ‘I could not have done otherwise, but it was up to me whether to raise my hand’. This judgement seems very odd, and even inconsistent. How could it have been up to me if I could not have done otherwise? If we leave it at that, we suspect that the statement expresses an irrational couple of beliefs. However, suppose that we do not leave it at that and further specify that I am a compatibilist philosopher. I believe a whole, further theory that also addresses opposing intuitions about the relevant concepts and explains why the conjunction sounds odd. Without filling in the relevant compatibilist theory of what ‘up to me’ means, we would not expect someone to hear the conjunction as something normal, or even coherent. In fact, that is what the philosophical theory of compatibilism is supposed to do: make sense of apparently conflicting claims. Furthermore, the compatibilist thinks that rejecting the possibility that the two conjuncts are compatible leads one into deeper philosophical trouble, and perhaps it leads one into endorsing an even more absurd-sounding conjunction. So, we should not take the fact that the isolated conjunction sounds odd, in conversation or in thought, as a good indication of whether compatibilism is a rational theory to hold. In general: when evaluating a position, we should not just pick two sentences from it and conjoin them to see how it sounds. We should consider the entire

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

74

Agnosticism

position or theory, especially when the rest of the theory contains an explanation of why the two conjuncts may sound odd together even though they are both true. Plausibly, this fits the case of the agnostic believer, or the view that agnostic belief is not always irrational. Agnostic belief may, and perhaps should, be accompanied by the claim that no attitude about p whatsoever is non-arbitrarily justified, since evidence does not bear on the issue at all, or it is outside the scope of epistemic standards. So, rather than consider just the conjunction, ‘I believe h and there is no evidence for h’, we should consider: I believe h and evidence is silent on h, so I have no reason to have any attitude whatsoever about it; h is the sort of proposition that is outside the scope of non-arbitrary epistemic reasons; from the perspective of accuracy, no attitude is more justified than any other about h, but nor is any attitude forbidden . . . Consider also that our practice of evaluating beliefs, and particularly evaluating them for support by evidence and overall coherence, usually takes place in everyday contexts in which evidence is at play and on the table. That is, evidence bears on the vast majority of beliefs that we usually evaluate. But agnostic belief is, explicitly, different. So, we should expect that our kneejerk reactions to the mere fact that one is agnostic and also believes are not as accurate as our careful consideration of the entire situation. In other words, we should reflect, reason, and do philosophy in order to figure out what is wrong, if anything, with agnostic belief when it is accompanied by an epistemological theory.

VII. CONCLUSION: WITH RESERVATIONS AND CONCESSIONS, AGNOSTIC BELIEF NEED NOT BE IRRATIONAL The question I set out to answer is: What is wrong with agnostic belief, from the epistemic perspective of accuracy? I considered what I take to be the three main ways to answer it by showing that it is always irrational to be an agnostic believer. But none of those answers succeeded, so I do not think they can explain what is wrong with agnostic belief. It seems, then, that agnostic belief might be defensible after all. This does not imply, of course, that all agnostic beliefs are

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

75

defensible. My aim in this section is to consider what features a defensible agnostic belief must have. To do this, I discuss the concessions made above in defending agnostic belief from the various challenges posed by the three attempted answers, as well as some additional issues that arise about agnostic belief.

(a) On the Partial, or Non-categorical Nature of the Agnostic’s Belief In order to defend agnostic belief against the various arguments considered above, I have suggested that it cannot involve a categorical belief. The content, h, cannot be used in an unqualified way in one’s reasoning or plainly asserted in communication in all contexts. It is, in other words, not anything like a certainty for the believer. Instead, agnostic belief in h should be a state of confidence in h, where one ‘thinks’ that h, or perhaps one is pretty sure, but not certain, that h. The agnostic believer expects, to some degree, that h is true, but not unreservedly or unqualifiedly. What, then, can the agnostic believer assert, to herself in thinking or others in communication, if not the content of her belief, h? As a believer, the agnostic believer does not hold merely that it might be that h is true; she should be willing to assert more than just ‘maybe h’. A more confident agnostic believer could assert ‘I think that h’, ‘I am pretty sure that h’, or even ‘probably h’. But a more conservative agnostic believer should stick with more reserved expressions such as ‘it may well be that h’. In all of these cases, the possibility of h has not been foreclosed and is thought to be more plausible than not-h. Another noteworthy aspect of agnostic belief ’s non-categorical nature is that it may still feel categorical, or practically certain, to the subject. Agnostic belief may feel like a practical certainty because inquiry into h is closed, and this is a central characteristic of categorical belief as Owens and Huemer characterize it. But in the case of h, inquiry and evidence are off the table by default, because there can be no such evidence according to her agnosticism. In other words, the case is closed because of the content of the belief (or the nature of the hypothesis), not because of the attitude that the subject has towards the content. So while it may feel like some sort of practical certainty, defensible agnostic belief in h is not categorical belief, since it is held in conjunction with recognition that h may be false.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

76

Agnosticism

That agnostic belief involves the closure of investigation, and thus feels like categorical belief, distinguishes it from some other states that one may have associated with agnosticism. For instance, Friedman35 has recently offered both a compelling case against reducing agnosticism to a degree of confidence—which is completely independent of any of the considerations I have offered—and a positive suggestion about what attitude agnosticism consists in. Her positive suggestion is that being agnostic is, roughly, keeping an inquiry open. To be agnostic is to consider it unsettled whether h, and therefore to continue to inquire about it. However, I doubt that openness, or inquiry, are necessary ingredients in agnosticism. I have described a type of agnosticism in which the agnostic takes there to be no foreseeable way to get evidence about h, so further inquiry would be futile. Inquiry, at least in the sense of consulting one’s sources of evidence, into h is closed for my agnostic. So even if we take continuing an inquiry about something to be an available doxastic state in general, this state does not seem to be justified by, let alone an essential part of, agnosticism. Perhaps other facts about one’s situation, over and above the facts that justify second-order agnosticism, make it reasonable to open some other sort of inquiry. For example, if you think that no known sources of evidence are available only because our concepts are underdeveloped here, then you may attempt a conceptual shift. That is a sort of inquiry. But notice two things: first, this is not essential to the sort of agnosticism I described, rather it is something that could happen in some such cases. Second, this is not a straightforward search for evidence. It is, in effect, a search for new kinds of evidence, or a new kind of question to investigate (since the concepts expressed in the question are being changed).

(b) What Else Should the Agnostic Believer Believe and Do? The agnostic believer should hold some further attitudes about the status of her beliefs, aside from merely that the belief is not justified by evidence. The agnostic believer should think that epistemic norms, or

35

Friedman (2013).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

77

rationality, have nothing non-arbitrary to say about whether to believe h. If instead one regards the lack of evidence for h as a good, significant reason to withhold belief in h (say, because one is not one of James’ adventurous types), then it would indeed seem incoherent, or inconsistent, for one to go on and believe h anyway. The agnostic believer need not regard her belief in h as positively justified, but she should certainly not regard her own belief in h as forbidden, or a bad thing to believe from an epistemic perspective. In other words, the agnostic believer must regard the significance of her lack of evidence for h as no indication at all about what to think about h. As I noted above, non-categorical belief presumably explains and causes actions. Is there anything distinctive about the way agnostic belief explains and causes actions, or should? Given that the agnostic believer holds that evidence does not support her belief, it seems right for her to exercise caution and humility when acting on her belief, especially when the action affects others’ well-being. And, when expressing her belief in h, it seems she should specify (if the context is appropriate) that the evidence does not support it. How can an agnostic believer take her belief seriously as a premise in her own thinking, given that she aims at accuracy and she knows that there is no indication that h is true? This has been addressed above to some extent, when I suggested that she should not use the premise that h in her reasoning (or in thinking to herself) without qualification in every context. Her thinking to herself that h, when she is in the process of reasoning to some other, descriptive or factual conclusion, should be framed, either explicitly or in the way it is used in the context, under the proviso that she has no reliable indication that h is true. This should insulate her from concluding anything much about the world on the basis of h, with any confidence, solely on the basis of h. In other words, her confidence that h is true is entirely epistemologically detached from her other beliefs about the world, cut off from her inferences about the here and now. This is not an overly restrictive constraint, because, by hypothesis, the content of her belief, h, has no relevant informational connection to things about which we can get evidence, or else it is not a candidate for secondorder agnosticism (see section I). So it is not clear how an inference could take one from h to conclusions that can ever be verified anyway. This leaves open, however, that the agnostic believer’s confidence in h stimulates non-doxastic states such as wondering about h, and perhaps some affective attitudes arising from whatever the details of h

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

78

Agnosticism

are. Perhaps some things are reasonable to hope for only given h, so those hopes are enabled by agnostic belief. If two agnostics disagree about whether h is true—that is, one of them is an agnostic believer and the other is an agnostic non-believer—what is the upshot of the disagreement? What do they learn from each other in their disagreement? It seems that neither of them learns about the probability of h, since both think that evidence for h is absent. Rather, each one learns something about the other’s doxastic policies or inclinations. The non-believer, for example, learns that the believer is an adventurous type when it comes to such topics, and this may be valuable information about a member of one’s epistemic community. It seems important, then, for the agnostic believer to present her belief as it is, unsupported by evidence, when engaging in debate. As for basing decisions that primarily affect one’s own life, it seems the consequences of agnostic belief can be both intellectually and emotionally significant. How could acting on h produce some significant and noticeable effect, given how removed h must be from the problems and issues encountered in this world? H, after all, cannot be indicated or implicated in any way by anything one experiences or foresees experiencing, or else it is not a candidate for justified secondorder agnosticism (see section I). But the agnostic believer may well spend more time wondering, for example, what the transcendent realm is like, or how this affects the significance of her current life. Perhaps this results in new affective attitudes towards this life that she would otherwise lack. To see this, consider one last example. Consider an ‘h’ that concerns a transcendent reality in which we somehow live on, beyond this life. That you think h might well be the case would explain your wondering about what this further life might be like, what the relation between your current life and this further realm is, and so on. Schellenberg36 has even suggested that certain forms of ritual are made rational by such ‘wondering’ or imaginative attitudes befitting an agnostic or ‘sceptic’ (whether or not the agnostic believes). It seems to me that these activities are not philosophically trivial. Even if such reflections and attitudes turn out to be a-rational, or outside the scope of the sorts of reasons and evidence that this life

36

Schellenberg (2009).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

What is Wrong with Agnostic Belief?

79

provides, they are the sort of reflection that an agnostic believer can engage in without any incoherence or blatant irrationality.

BI B LI OGR APHY Avnur, Yuval. ‘Mere Faith and Entitlement’. Synthese 189, no. 2 (2012a), pp. 297–315. Avnur, Yuval. ‘In Defence of Secular Belief ’. In Oxford Studies in Philosophy of Religion, edited by Jonathan Kvanvid (Oxford: Oxford University Press, 2012b). Avnur, Yuval. ‘Denial, Silence, and Openness’. The Meaning and Power of Negativity, edited by Ingolf Dalferth (Tübingen: Mohr Siebeck, forthcoming). Chignell, Andrew. ‘Belief in Kant’. Philosophical Review 116, no. 3 (2007), pp. 323–60. Chislenko, Eugene. ‘Moore’s Paradox and Akratic Belief ’. Philosophy and Phenomenological Research 91, no. 3 (2014). Christensen, David. Putting Logic in its Place: Formal Constraints on Rational Belief (Oxford: Oxford University Press, 2004). Coliva, Annalisa. ‘How to Commit Moore’s Paradox. Journal of Philosophy 112, no. 4 (2015), pp. 169–92. Friedman, Jane. ‘Rational Agnosticism and Degrees of Belief ’. Oxford Studies in Epistemology 4 (2013), pp. 57–81. Gendler, Tamar Szabó. ‘Alief and Belief ’. Journal of Philosophy, 105, no. 10 (2008), pp. 634–63. Gleiser, Marcello. The Island of Knowledge (New York: Basic Books, 2015). Goldman, Alvin I. (1979) ‘What is Justified Belief?’, Epistemology: An Anthology, edited by Ernest Sosa and Jaegwon Kim (Oxford: Blackwell, 1979), pp. 340–53. Hawthorne, John; Rothschild, Daniel, and Levi Spectre. ‘Belief is Weak’. Philosophical Studies 173, no. 5 (2016), pp. 1393–404. Horowitz, Sophie. ‘Epistemic Akrasia’. Noûs 48, no. 4 (2014), pp. 718–44. Huemer, Michael. ‘Moore’s Paradox and the Norm of Belief ’. In Themes From G. E. Moore: New Essays in Epistemology and Ethics, edited by Susana Nuccetelli and Gary Seay (Oxford: Clarendon Press, 2007). Huxley, Thomas Henry. ‘Agnosticism: A Symposium’. In The Agnostic Annual, edited by Charles Watts (1884), pp. 5–6. James, William. The Will to Believe and Other Essays in Popular Philosophy (New York: Longmans, Green, and Co., 1897 [1896]). Lackey, Jennifer. ‘Norms of Assertion’. Noûs 41, no. 4 (2007), pp. 594–626. Littlejohn, Clayton. ‘Moore’s Paradox and Epistemic Norms’. Australasian Journal of Philosophy 88, no. 1 (2010), pp. 79–100.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

80

Agnosticism

Monton, Bradley. ‘Bayesian Agnosticism and Constructive Empiricism’. Analysis 58, no. 3 (1998), pp. 207–12. Moran, Richard A. Authority and Estrangement: An Essay on Self-Knowledge (Princeton: Princeton University Press, 2001). Owens, David. Reason without Freedom: The Problem of Epistemic Normativity (London: Routledge, 2000). Plantinga, Alvin. Warranted Christian Belief (New York: Oxford University Press, 2000). Ribeiro, Brian. ‘Epistemic Akrasia’. International Journal for the Study of Skepticism 1, no. 1 (2011), pp. 18–25. Rosenkranz, Sven. ‘Agnosticism as a Third Stance’. Mind 116, no. 461 (2007), pp. 55–104. Russell, Bertrand. ‘Is There a God?’. The Collected Papers of Bertrand Russell, Vol. 11: Last Philosophical Testament, 1943–68 (London: Routledge, 1952), pp. 547–8. Schellenberg, J.L. Prolegomena to a Philosophy of Religion (Ithaca, NY: Cornell University Press, 2005). Schellenberg, J.L. The Will to Imagine: A Justification of Skeptical Religion (Ithaca, NY: Cornell University Press, 2009). Schellenberg, J.L. Evolutionary Religion (Oxford: Oxford University Press, 2013). Schellenberg, J.L. The Hiddenness Argument: Philosophy’s New Challenge to Belief in God (Oxford: Oxford University Press, 2015). Schellenberg, J.L. ‘Three Ways to Improve Religious Epistemology’. Philosophy (forthcoming). Sober, Elliott. ‘Absence of Evidence and Evidence of Absence: Evidential Transitivity in Connection with Fossils, Fishing, Fine-tuning, and Firing squads’. Philosophical Studies 143, no. 1 (2009a), pp. 63–90. Sober, Elliott. ‘Parsimony Arguments in Science and Philosophy—A Test Case for Naturalism’. Proceedings and Addresses of the American Philosophical Association 83, no. 2 (2009b), pp. 117–55. Unger, Peter K. Philosophical Relativity (Oxford: Oxford University Press, 1984). Van Leeuwen, Neil. ‘Religious Credence is not Factual Belief ’. Cognition 133, no. 3 (2014), pp. 698–715. White, Roger. ‘Problems for Dogmatism’. Philosophical Studies 131, no. 3 (2006), pp. 525–57. Williamson, Timothy. Knowledge and its Limits (Oxford: Oxford University Press, 2000). Wright, Crispin. ‘Warrant for Nothing?’ Aristotelian Society Supplementary Volume 78, no. 1 (2004), pp. 167–212.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

3 If Agnosticism, Then What? Francis Fallon

I. INTRODUCTION What follows1 will not argue directly for religious agnosticism. Let us instead take agnosticism, to appropriate William James’ phrase, as a ‘living option’. This chapter will aim to elaborate the implications of agnosticism. If we stipulate agnosticism, what attitudes and behaviours count as rational? If this leads to a qualified acceptance of a broad range of practices, this may indirectly support agnosticism, at least to some of those who might otherwise have conceived of agnosticism more narrowly. So, agnosticism here is antecedent, the premise; what, then, are the consequences? On one side of this question, we find Clifford, who advocates a strict rejection of belief in the face of uncertainty: ‘To sum up: it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence’. A sense of urgency motivates such a bold claim, for Clifford believes ‘Every rustic who delivers in the village alehouse his slow, infrequent sentences, may help to kill or keep alive the fatal superstitions which clog his race.’2 If agnostic, then don’t believe. On the other side lie responses that argue that, in the absence of certain knowledge of God’s existence, belief is nonetheless the most rational attitude, or perhaps even compelled on ‘supra-rational’ grounds. In the former case, the epistemological requirement for belief is less demanding than Clifford maintains. On the latter (often associated with 1 2

My gratitude to an anonymous reviewer for several helpful comments. Clifford (1886 [1877]).

Francis Fallon, If Agnosticism, Then What? In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0004

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

82

Agnosticism

Kierkegaard), extra-epistemological concerns determine our attitudes in cases of uncertainty. Both types may fairly, if not quite thoroughly, be labelled ‘leap of faith’ claims. If agnostic, then believe. Not every attitude regarding God affords characterization by reference to belief. Some (following Wittgenstein) regard faith not as a matter of belief in God’s existence, but as a ‘form of life’. Such a line is orthogonal to the current discussion. No doubt it plausibly characterizes how some incorporate religion into their lives. Nevertheless, many explicitly practice religion as a reflection of their theistic beliefs. Moreover, the ‘forms of life’ interpretation does not speak directly to the rational implications of accepting agnosticism. Even if religion is a form of life, God, understood as a cause or sustainer of being, may or may not exist. On Wittgenstein’s gloss, the attitudes and practices appropriate to the agnostic remain unexplicated. Another approach that sidesteps the issues normally associated with belief involves alternative definitions of God, that is, not defining God as a creator or sustainer of being. Some define God very broadly. Spinoza, for example, identifies God with nature. More recently, Cupitt draws on the Christian tradition in order to identify God with love.3 In either case, to define God so broadly obviously almost entirely alters the discussion. Apophaticism alters the discussion as radically, although in a different way: it holds that God eludes positive definition altogether: we can only describe what God is not. Again, whatever the merits of these approaches, they do not reflect how many—whether theist, atheist, or agnostic—address the question of God. If God is nature, existence, or love, then in some strained sense this prescribes that we all adopt theism, although the implications for attitudes and practices beyond this remain unclear. Apophaticism shares with agnosticism a principled non-committalism.4 This in one sense makes it a form of agnosticism, but is still different from what perhaps is more typically meant by the term: Apophaticism denies that we can describe God, whereas typical agnosticism declines to accept or reject the existence of God. The concern of this chapter, to trace the implications of agnosticism about a creator or sustainer God, builds into its premise a conception of 3

Cupitt (2008). This is not to say that all forms of apophaticism are totally non-committal. Some apophatic theists, for example, believe that while we cannot know God’s intrinsic properties, God may be known to exist in virtue of God’s relation to us as saviour or creator. 4

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

83

religion different from, or at least not fully covered in, Wittgenstein’s approach, and a conception of God not identical to the God of Spinoza, Cupitt, or apophaticism. Adherents of any of these philosophies of religion will find no argument directly contradicting them in what follows, but an argument that begins with different assumptions. To anyone who conceives of God as a creator or sustainer, and is unsure of that God’s existence, the argument may be a matter of more interest. The argument will take the following shape: section II will situate agnosticism philosophically with respect to atheism and theism, entertaining epistemologically agnostic positions that nevertheless prescribe belief. Agnosticism, as envisioned here, does not accept certain arguments against theism as conclusive. Although it is taken as a premise that theism is not demonstrated to be true, some make the case that belief in God is justified on grounds other than such demonstration. This section addresses several such related arguments. These ultimately do not justify sustained belief in God’s existence, but their merits help in characterizing the agnostic position as one of hopeful ambivalence concerning God’s existence. Section III explores this notion of ambivalence: section III(a) discusses the question of God as a mystery, and offers a metaphor for the agnostic as detective. Section III(b) identifies the life and work of Oscar Wilde as exemplifying an active form of ambivalence; this section entertains his metaphors of agnostic as wanderer and, more positively, agnostic as engaged in flirtation. Section III(c) stakes out epistemological ground for openness to episodes of religious experience within an overarching agnostic framework. This includes the introduction of one argument from analogy and the rejection of another. Section IV, building on the philosophical characterization of agnosticism from section II and the exploration of ambivalence in section III, addresses more directly what attitudes and practices are rationally justified for the agnostic. This section argues that a very broad range of both religious and non-religious lifestyles fall within the rational scope of agnosticism.

II. RELATING AGNOSTICISM TO THEISM AND ATHEISM Since we are here concerned with rational accounts of the (non-) existence of God, the most relevant form of theism involves claiming

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

84

Agnosticism

that there is a rational argument for concluding that God exists. By definition, agnosticism holds that God might exist. So, it follows that the theistic conclusion might be correct. At the same time, the agnostic denies that there is a known, rationally compelling argument for this conclusion. Such an argument might exist, but the agnostic has either never encountered it or has failed to appreciate its soundness. These claims leave open whether religious experience or revelation is genuinely of God. For all the agnostic knows, the theist inspired by experience has access to the truth of God’s existence. This experience might constitute a privately compelling reason for faith, but not one that transfers, that is, it is not rationally demonstrable to those without such access. (This agnosticism is more expansive than the ‘permanent agnosticism’, described in the Introduction to this volume; for our purposes, the relevant agnostic takes it as possible that compelling evidence for God’s existence may yet come to light.) Agnosticism, of course, is free to reject anti-theistic arguments. For example, Russell argues against theism via analogy.5 Were he to posit the existence of a teapot revolving in orbit, it should be his burden to prove its existence, and not the burden of others to disprove it. The cases are not analogous, though. The teapot plays no explanatory role in anything, and the existence of God might.6 In addition, the burden of proof often is assumed by theistic arguments, such as ‘wager’ arguments (of which more below).

(a) Compelling Grounds for Belief, Even Given Epistemological Agnosticism? Plantinga’s Basic Beliefs and Gutting’s Multiple Dimensions One kind of approach begins by accepting epistemological agnosticism concerning God, but argues that despite this uncertainty, there still exist compelling grounds for belief in God’s existence. Plantinga may be read as offering one such account. He claims that some beliefs are properly basic and do not stand in need of general justification, and that the belief in God is one such belief.7 Does this not permit us to believe incredible things? Exploring this possible objection to his 5

Russell, Feinberg, and Kasrils (1969). See Garvey (2010) for an extended discussion of Russell’s teapot; see also the Introduction to this book. 7 Plantinga (2006 [1980]), pp. 58–60. 6

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

85

position, Plantinga entertains an objection similar to Russell’s teapot. Plantinga uses the example of the character Linus, from the Peanuts comic strip (of Charlie Brown fame). Linus believes that each year at Halloween, the Great Pumpkin visits the most sincere pumpkin patch, and at midnight rises from it, flying through the air distributing presents. What could distinguish belief in God as properly basic without admitting belief in the Great Pumpkin as possibly properly basic? Plantinga responds by claiming that basicality is only identifiable on a case-by-case basis. One’s belief in the existence of other minds, or that one had breakfast this morning, finds justification not by satisfying necessary and sufficient conditions ‘from above’. Rather, the justification depends on inductive evidence ‘from below’. In other words, the justification comes not from abstract argumentation, but from particular contexts. The existence of God, unlike that of a Great Pumpkin, and still less like that of a teapot between Earth and Mars, would offer a framework for experience, even a meaning for the narrative of life. Considering the following verses from Wordsworth may help to clarify what Plantinga means by evidence of God’s existence ‘from below’, that is, from particular contexts: A presence that disturbs me with the joy Of elevated thoughts; a sense sublime Of something far more deeply interfused, Whose dwelling is the light of setting suns, And the round ocean and the living air, And the blue sky, and in the mind of man; A motion and a spirit, that impels All thinking things, all objects of all thought, And rolls through all things.8

We may read these as illustrating how a belief in God originates from particular contexts, as Plantinga describes. God dwells in the particulars of life—here starting with setting suns and ending in the mind of man. This is not revelation, but might be described by Calvin’s phrase, ‘sensus divinitatis’. Many theists give something along these lines as the basis of their faith.

8 Wordsworth (1798). My thanks to Mollie Smith for suggesting this relevant passage.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

86

Agnosticism

In subsequent work, Plantinga has expanded on this. The ‘Aquinas-Calvin’ (AC) model takes it to be the case that sensus divinitatis is a properly functioning cognitive feature, an apt reception of experiences. Not all basic beliefs are equal, and some become exposed as false. Beliefs in the Great Pumpkin or voodoo meet this fate, but theistic beliefs do not.9 Plantinga does not argue that this model or the basic beliefs in God that it includes are true, but rather that, if it is true that God exists, the model is likely true, and therefore that belief in the model, and the model’s basic beliefs, would be warranted. ‘The source of warrant’ for theism is still ‘not argument of any sort’.10 Plantinga does argue that the model is possible, and ‘subject to no philosophical objections that do not assume that [theism] is false’.11 The agnostic certainly can accept the former, but may reject the latter.12 That is, the agnostic position, at least as we have defined it, may accept that there are objections that block belief in God’s existence from being rational. Moreover, sensus divinitatis is susceptible to non-theistic explanation, as discussed in much contemporary work. This fact alone does not disprove God’s existence, but certainly does not help the theistic cause. Plantinga’s case does not salvage belief in God from agnosticism about God’s existence. Indeed, as Plantinga himself writes: ‘But is [the AC model] true? And here we pass beyond the competence of philosophy, whose main competence, in this area, is to clear away certain objections, impedances, and obstacles to [theistic] belief ’.13 Anyone, however—the agnostic included—can accept that the theistic position just outlined is more plausible than claims about teapots and Great Pumpkins. Moreover, this version of theism has a better claim to rational transferability than arguments from authority or personal direct revelation: many who do not take themselves to have experienced God directly find some satisfaction or appeal in framing their experience of the world by reference to a present God who ‘rolls through all things’. Still, the agnostic takes the position that this account may be true, but also may be only an articulation of the psychology of one variety of religious experience.14 9

10 Plantinga (2000), p. 350. Plantinga (2000), p. 352. Plantinga (2000), p. 351. 12 See Plantinga (2000), pp. 357–499 for Plantinga’s considerations of various possible ‘defeater’ arguments, i.e. attempts to argue that theistic belief is irrational. 13 Plantinga (2000), p. 499. 14 See Grayling (2008), p. 2, for a non-theistic gloss. 11

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

87

Melville conveys at least a similar attitude, a non-committal sympathy for a religious worldview, when he writes in Moby Dick: And so, through all the thick mists of the dim doubts in my mind, divine intuitions now and then shoot, enkindling my fog with a heavenly ray. And for this I thank God; for all have doubts; many deny; but doubts or denials, few along with them, have intuitions. Doubts of all things earthly, and intuitions of some things heavenly; this combination makes neither believer nor infidel, but makes a man who regards them both with equal eye.15

But is this satisfactory? Later in the novel, Melville describes the restlessness of agnosticism: There is no steady unretracing progress in this life; we do not advance through fixed gradations, and at the last one pause:–through infancy’s unconscious spell, boyhood’s thoughtless faith, adolescence’ doubt (the common doom), then skepticism, then disbelief, resting at last in manhood’s pondering repose of If. But once gone through, we trace the round again; and are infants, boys, and men, and Ifs eternally. Where lies the final harbor, where we unmoor no more?16

It is (probably deliberately) ambiguous whether this question should be read as rhetorical. Even if Plantinga’s case does not ground theism, perhaps there is another way to move from epistemological agnosticism to a less restless, more stable position of faith. Gutting describes ‘multidimensional’ arguments for the existence of God as broadly inductive, in a way that echoes Plantinga. One kind of multidimensional argument takes experiences as ‘delivering’ the conclusion directly. In the other, experiences—as well as other possible evidence from argument—give rise to an explanatory hypothesis. This constitutes an inductive claim because the evidence does not necessitate the theistic conclusion. Rather, it ‘puts us in a position to judge that the conclusion is true, without logically forcing the judgment’.17 Certain experiences inspire us to religious thought or feeling, and questions concerning the meaning of existence find incomplete or obscured but relevant responses on a religious worldview; perhaps not separately, but together, these are enough to justify religious belief.

15 17

16 Melville (1998 [1851]), pp. 383–4. Melville (1998 [1851]), p. 501. Gutting (1982), p. 110; emphasis in original.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

88

Agnosticism

Gutting himself, though, says that these are not compelling, but ‘put us in a position to judge the conclusion as true’. This is not the uncontroversial position that the arguments are not deductively sound. Successful inductive and abductive arguments can nonetheless be compelling. Gutting is not suggesting that the arguments on offer here for God’s existence have the same status as those for other minds, or for the conclusion that the sun will rise tomorrow morning. Rather, it seems like he is claiming that the arguments are enough to raise the belief in God’s existence to the (more modest) status of reasonable option. But if the reasons aren’t rationally compelling (even in the weaker, in/abductive sense), then why not judge that the conclusion may be true, as does the agnostic? Isn’t this, rather than the theistic position, the justified account? These, of course, are not rhetorical questions. Neither Plantinga’s account of basic beliefs nor Gutting’s related multidimensional account suffice as reasons to prefer a theistic leap to an agnostic refrain from (dis)belief, but perhaps further justification is available.

(b) Wager Arguments: Pascal, James, and Gutting Pascal famously presents his argument for belief in God as a wager, although this gives the misleading impression that he takes faith to a matter of a simple decision.18 In fact, he regards it as a very serious responsibility that demands deliberation and practice. The title of James’ ‘The Will to Believe’ could aptly characterize Pascal’s claims. Both are wager arguments, and both give reasons for willing ourselves to believe something for which we have insufficient evidence. Pascal holds that we must either believe or disbelieve in God’s existence. We must place a wager either way: ‘It is not optional. You are embarked.’ Pascal’s fragment does not give a clear, explicit reason for this dichotomy, but the very same idea motivates James’ argument, and he does elaborate. In one place, the claim is psychological: ‘Our passional nature not only lawfully may, but must, decide an option between propositions.’19 Now, perhaps the psychological nature of some people denies them the possibility of negotiating a 18

Pascal (2016), p. 51. It is worth noting that Pascal’s own faith was not dependent upon this argument, and that the argument itself is a posthumously published fragment. 19 Pascal (2016), p. 11; emphasis in original.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

89

position between theism and atheism, but this certainly does not hold generally. That is, not all people find themselves psychologically unable to refrain from choosing between the two poles of atheism and theism. Those who do are at liberty to assess for themselves the implications of the psychological dilemma, and may choose to follow James’ argument to will belief. For the rest of us, however, this particular claim does not rule out the rationality of simultaneously declining either to believe or disbelieve in God’s existence. Elsewhere in his essay, James describes the same dilemma somewhat differently, opening up the possibility that the issue is not merely psychological: ‘there is really no scientific or other method by which men can steer safely between the opposite dangers of believing too little or of believing too much’.20 James, like Pascal, never quite gives a positive argument for this rigid dichotomy. Let us bracket this question now, though, because if either succeeds in giving extra-epistemological justification for placing faith in God’s existence,21 the point is moot. Pascal claims that, given that God would reward the faithful with everlasting life, it makes sense to affirm the existence of God and practice faith, despite not having certain knowledge of God’s existence. Even theists can question whether God’s reward is contingent in this way. Pascal also claims that faith will be a boon to life even if God turns out not to exist. Likewise, James associates faithful living with an improved life whether God exists or not, emphasizing the point by ending his essay with a quotation from Fitz James Stephen: ‘If death ends all, we cannot meet death better’ (James (2009 [1912]), p. 22). Like the argument from possible everlasting life, this is an extraepistemological, specifically utilitarian, argument for belief. In its favour, it does not rely on a specific conception of the mind of God, but nevertheless the claim is weak. The correspondence between religiosity and improved living is undeniable in some cases, however one defines ‘improved living’. Still, (like James’ psychological claim) this certainly does not hold true generally. 20

James (2009 [1912]), Preface. We could characterize Pascal and James as arguing that we have an obligation to will belief in God’s existence, as the claim seems to be one of necessitation, more than mere justification, and both acknowledge that such belief might not be readily forthcoming psychologically. On this reading, James is understating his case when he describes it as ‘a defence of our right to adopt a believing attitude in religious matters’ James (2009 [1912]), p. 2. 21

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

90

Agnosticism

More recently, Gutting has offered an extra-epistemological argument for belief that differs from the earlier multidimensional accounts and refines upon the ‘wager to will’ arguments of Pascal and James. In a piece entitled ‘Pascal’s Wager 2.0’, Gutting abandons the claim that one should wilfully believe in God despite a lack of convincing evidence. He instead suggests a more moderate stance: one should ‘embrace a doubt of desire rather than a doubt of indifference’.22 This move is not very controversial. A creator God might be beneficent. The existence of a creator God could suggest the existence of an afterlife. It is clearly a rational attitude under such conceptions of God to prefer that God exists than not to care. (If one’s conception of a creator God takes that God to be unforgiving, of course, the implications change.) From here, though, Gutting argues for ‘religious agnosticism: serious involvement with religious teachings and practices’. So the prescription extends beyond a hopeful attitude to an engagement with doctrines and rituals of faith. Like earlier arguments, the possibility of ‘making appropriate contact with [a beneficent] power’ motivates the religious turn. Not only does Gutting’s wager differ by requiring something less than belief, it also takes into consideration that prior convictions, including ethical ones, will and should constrain one’s engagement with religion. Given this condition, Gutting’s wager avoids the problem of naively associating religiosity with improved living. Nor must the ‘serious involvement’ he recommends come at the cost of other ‘worldly satisfactions’: ‘we can decide for ourselves’. That is, the nature of the engagement with religion is left open enough that one need not (but is free to) forgo the pursuits that one values that fall outside of religion. For these reasons, Gutting contributes a much better account of what rationally follows from uncertainty about God’s existence. Still, it is by no means clear that ‘serious involvement’ with religion would be necessary for contact with God. In response, one might argue that it increases the likelihood of contact with divine. If the agnostic has reasons to accept the latter claim, then it would indeed count in favour of examining religious doctrines and participating in religious practices, but this perhaps assumes too detailed a conception of God’s nature on the part of the agnostic. In other words, perhaps the

22

Gutting (2015).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

91

conviction that God would reward religious engagement by making contact more reliably with those who practice it might not be wellfounded. Moreover, the disclaimers Gutting builds into his argument might stand in tension with his own characterization of it as compelling ‘serious involvement’. This could happen in at least one of two ways: First, one’s ethical commitments might require so much as not to allow time or mental energy for such involvement: it is plausible that a religious lifestyle could interfere with at least some people’s most valued projects. Second, ‘serious involvement’ might necessarily include the revision of one’s ethical commitments, which would contradict Gutting’s reassurance that they would act as prior constraints (at least without an explication of what precisely is meant by ‘constraint’). Instead of entailing this level of involvement, the argument can be read as compelling an ‘open mind’. This gloss deemphasizes the necessity for engaging with teaching and practice and takes as central that, in Gutting’s words: ‘religious agnosticism demands only that I reject atheism, which excludes the hope for something beyond the natural world knowable by science’. This reading of Gutting fits agnosticism well. It distinguishes it from atheism, and requires an open mind towards belief in God’s existence. It gives good extra-epistemological reason to hope for God’s existence. It urges the plausibility of the claim that engagement with teaching and practice will conduce to contact with God (should God exist). At the same time, it respects the prior claims of science and ethics, and that one’s best life, including the role of religious engagement within it, will ‘depend on individual inclinations and capacities’. We have come a long way for the articulation of what agnosticism rationally entails, and it proves—perhaps unsurprisingly!—to be a cautious synopsis of what it means to have an open mind. More, of course, remains to be said by way of elaboration.

III. AMBIVALENCE OVER INDIFFERENCE The argument so far opposes both a general positive disbelief in God and a general belief in God (whether on probabilistic or extraepistemological grounds). Gutting, as we have seen, criticizes an agnosticism of indifference, and with good reason. In its place, let

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

92

Agnosticism

us entertain an agnosticism of ambivalence. Instead of ‘lax’ or passive agnosticism, let us explore an active form of agnosticism.23 Metaphor plays an important role within religion, both in moving people towards faith and in expressing the faith that they hold. Again, within religion, metaphors have philosophical significance, often of a great subtlety. As we entertain an agnosticism of ambivalence, we will retain an openness to metaphor and analogy that may help to serve our consideration of religion from without.

(a) Mystery Here is a fair way of characterizing agnosticism: The agnostic believes that we do not have a solution to the mystery of whether God exists. Etymologically, ‘mystery’ has a religious history. According to The Compact Edition of the Oxford English Dictionary, its ancient Greek root referred to ‘secret religious ceremonies . . . which were allowed to be witnessed only by the initiated, who were sworn never to disclose their nature’.24 In the Bible, the term has a variety of meanings, including, in the Old Testament, ‘secret purpose or counsel’ and, later in the New Testament, a truth revealed by God. During the Medieval era, mystery plays enacted scenes from the life of Jesus, and the term took on non-theological significance too. In the Early Modern era, the term referred also to Christian rituals. The New Catholic Encyclopedia adds that Paul’s usage of the word, once thought to be a borrowing from the pagan realm, reflected its currency as a theological term in Hebrew. In any case they are analogous.25 Old Testament usage holds mysteries as unknowable even to the wise, and consequently improper subjects of investigation. Paul’s usage reflects this. Contemporary theology preserves this sense in various ways. Catholic revelation theology holds that God can reveal mysteries to people, which ‘even after their revelation, continue to exceed comprehension’.26 ‘Mystery’—which ‘even to 23 In coining the term ‘New Atheists’, Wolf (2006) defines them as calling upon ‘lax agnostics’ to ‘exorcise . . . curse of faith’. Incidentally, this seems like an instance of what Chesterton calls ‘ill-tempered’ agnosticism (Chesterton (1993 [1925]), p. 11); perhaps our active agnostic can aim to be well-tempered. 24 OED (1971 [1889]). Apparently from ‘muein’: ‘to close the eyes or lips’ or ‘initiate’. 25 26 Brown (2003), p. 79. Brown (2003); see also Dulles (2003), p. 82.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

93

faithful and educated Christians remains obscure by reasons of its sublimity’—became, in the nineteenth century, ‘a technical term in the Catholic theology of revelation’.27 At the time, ‘semirationalists’ argued that reason alone was sufficient for comprehending faith, but this never gained doctrinal approval, and Vatican I explicitly lays out that certain truths of faith are properly mysterious and defy rational comprehension, although diligent reason affords some profitable understanding.28 Various Christian masses include the words ‘mysteries of faith’ (mysterium fidei), usually during or immediately following consecration and before the memorial acclamation. Tracing the religious roots of ‘mystery’ reminds us how close theistic approaches (including Jewish, Christian, and pagan varieties) may come to agnosticism. Theism can positively build unknowability into belief (something noted briefly in the earlier mention of apophaticism). Characterizing religion as ancient and science as modern can be unhelpfully vague and, if taken generally, misleading or even false;29 in the context of elaborating attitudes to mystery, though, it allows us to identify a relevant distinction. The religious incorporation of mystery might appear ad hoc, but its ancient roots explain the motivation. Before the Scientific Revolution, the project of gaining knowledge concerned the right interpretation of religious and authoritative ancient texts. Mystery, conceived of as impervious to future investigation and speculation, makes sense in an intellectual setting that appealed to tradition for answers. Tradition conferred authority on the essential mysteriousness of aspects of God, and religion continues to accept this. Discovery, understood as new 27 Dulles (2003), p. 83. In the second half of the nineteenth century, ‘prominent [Anglican] theologians were calling doubt “highly useful” and a necessary “borderland between knowledge and ignorance” ’ (Lane (2011), p. 7). 28 John Paul II’s (1998) papal encyclical, Fides et Ratio, emphasizes the combination of reason and faith as necessary to reaching the truth of God’s existence, but even here reason alone will not suffice. (This is not, of course, to deny that there is a long philosophical tradition –stretching back further than Aquinas but taking on new momentum in the seventeenth century—that takes the demonstrability of God’s existence as a requirement of faith: See Buckley (1987), pp. 79–85. Nor is it to deny that the Catholic Church takes God’s existence as demonstrable. Both Vatican I and Vatican II expressly state that God’s existence is rationally demonstrable, while still categorising various claims about God’s nature as mysterious. (The Catholic Church has not emphasized this to its uneducated members: see (Kors [1990], pp. 111–18).) I owe this point to discussion with Gavin Hyman.) 29 Langer (1979 [1942]) is an exception.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

94

Agnosticism

knowledge, only receives widespread reification, currency, and respectability during the Scientific Revolution.30 Rationality has a place for according tradition due respect and attention, but it also requires an open search for knowledge. We can extend our employment of the term ‘mystery’ in a useful way: The language of mystery, especially in a modern context, lends itself to the metaphor of detection. Theist and atheist alike regard the case of God’s existence as closed. Theism and agnosticism differ on how to interpret the mystery of God. The agnostic (or at least the open-minded agnostic I have in mind for this essay) claims that the case remains open. When confronted with a mystery, it is permissible to entertain suspicions. Indeed, a detective has no alternative but to entertain suspicions: it is part of the job description. The rationality of this general point applies to the particular case of God. In the face of an unsolved mystery concerning God’s existence, and without any access to God through revelation, the rational person should entertain suspicions, but remain open or ambivalent. How far can this detective metaphor take us, though? A real detective assesses empirical evidence and finds the most compelling theoretical fit. Such investigation could conclude with a true smoking gun, or a convincing confession. But the agnostic has weighed the evidence for God and found it wanting (to borrow language from the book of Daniel). Plantinga’s basic belief account and Gutting’s (1982) related multidimensional account fail to compel as theoretical glosses. They fit some of the evidence, but parsimony (not to mention arguments from evil or suffering) counts against accepting such interpretations. Furthermore, we cannot settle the question by appeal to extraepistemological concerns, as the treatment of Pascal, James, and Gutting (2015) shows. The case remains open, but the trail seems to have gone cold, and this might suggest that an attitude of indifference is appropriate after all. Actively keeping an open mind and reaching an impasse concerning God’s existence seem to stand in tension with one another. If, upon reflection, one finds compelling reason for neither theism nor agnosticism, and has no experience (as) of divine revelation or access—in other words, if even the epistemically responsible person

30

Wootton (2015).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

95

runs out of ‘leads’ concerning the existence of God—then doesn’t the investigation stall? Of course, one could deny the likelihood that even an epistemically responsible person could exhaustively examine the relevant body of literature and argumentation. This is probably true: some academics devote entire careers to becoming experts on (aspects of), for example, the work of Augustine or Aquinas. Secondary literature offers a certain kind of access, and a competent layperson can cover extensive ground, but this does not permit us to eliminate the possibility that a rational demonstration of either theism or atheism is either extant already or at least possible. Nevertheless, other epistemic issues require our attention. A detective will redirect his or her focus to other cases if they afford more fruitful investigation or are more pressing. But this is not indifference to the original case, and a good detective will in principle stay alert for, or remain receptive to, possibly relevant evidence from any sources. To appreciate what remains available to the receptive agnostic, we will turn to an example from intellectual history.

(b) Case Study: Oscar Wilde We typically read Wilde as glib and flippant, according to his design. These charges apply to his characters, not what they voice, and apply far less reliably to Wilde’s non-fiction, or even his life choices. Wilde’s protean religious position illustrates the range of attitudes and behaviours available under agnosticism.31 As a young man, Wilde, a Protestant by birth, actively entertained the idea of converting to Catholicism. As an undergraduate at Trinity College Dublin, he had studied the classics with J.P. Mahaffy. At Oxford, his interest in Catholicism grew. Between terms (in 1877), Mahaffy brought Wilde with him on a trip to Greece, in the express hopes of discouraging both this growing religiosity and Wilde’s related interest in travelling to Rome; Mahaffy evidently thought it 31 Although we have here focused on Wilde, there are other important cases of active, ambivalent agnosticism. According to Brown (1986), these would include the case of Darwin: ‘scholars have failed to appreciate fully the extent and importance of the ambivalence that characterized Darwin’s theology at every stage of its evolution, and particularly in its final phase of an intermittent and largely “agnostic” theism’ (p. 3, emphasis added; see also pp. 27–33).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

96

Agnosticism

necessary to encourage Wilde to be a ‘good pagan’.32 Whatever the influence upon Wilde, it did not extend to eliminating his attraction to religion altogether. At Oxford in 1878, Wilde sought counsel from the Catholic priest H.S. Bowden, and they arranged for Wilde to enter the Catholic faith. On the day of the ceremony, Wilde did not show, instead sending a bunch of lilies to the church.33 Wilde’s youthful writings reflect this ambivalence. In a letter (1876) to William Ward, he ‘confess[es] not to be worshipper at the Temple of Reason’, describing reason as ‘misleading and thwarting’ and faith as a ‘bright lantern for the feet, though of course an exotic plant in man’s mind, and requiring continual cultivation’. He remarks upon ‘the beauty and necessity for the incarnation of God into man to help us grasp at the skirts of the infinite’, before closing the letter with a note about reading Thomas a Kempis before bed.34 In his ‘Notebook Kept at Oxford’ (1874‒9), he seems much more dubious: ‘All religions are to the vulgar equally true, to the philosopher equally false, to the statesman equally useful’.35 Still, in the same text, he indicates a dissatisfaction with nonreligious explanation, perhaps conducive to a receptivity to religion: Matter[,] motion and force are merely symbols of the unknown cause or reality which is coextensive with all the phenomena of the world— Progress is as much to the unknown and unknowable as to the known; ultimately all accountable and natural facts are unaccountable and unnatural.36

In his roughly contemporaneous ‘Commonplace Book’, he expresses pessimism about the place of religion, rather vividly: Primitive religions contain the germ of philosophy and of physical science—unnatural children who seek to annihilate their mother when they have attained to their maturity[;] yet the intellectual synthesis between religion and science as it is commonly given to us is nothing more than a monstrous OEdipean union of vigorous manhood with the effete mother who bare it—a union whose children must be wanderers and born to evil things.37

32 34 35 36 37

33 Arkins (1999). Ellmann (1988), p. 94. In Holland (2006), pp. 18–19; Quintus (1991), p. 526, also quotes this in part. Smith II and Helfand (1989), p. 172; p. 102 in original ‘Notebook’. Smith II and Helfand (1989), pp. 160–1; p. 36 in original ‘Notebook’. Smith II and Helfand (1989), p. 124; p. 81 in original ‘Commonplace Book’.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

97

This description of children of science and religion as ‘wanderers’ lends itself to characterization as agnostic quite naturally. We might regard Wilde himself as an agnostic wanderer. As Quintus notes, in later works—‘The Soul of Man Under Socialism’ (1891), and ‘De Profundis’ (written in 1897)—Wilde ‘envisions the benefits that one can reach through Christianity as culturally remedial’.38 Still, during the same period, ‘when asked . . . what his religion was, Wilde replied, “I don’t think I have any. I am an Irish Protestant”’ (Ellmann 1988, pp. 94‒5).39 To borrow Melville’s words, Wilde’s religious life was ‘unmoored’. Controversy surrounds whether Wilde found a ‘final harbor’. On his deathbed, Wilde asked for a Catholic priest. His friends weighed his solemn visits to Catholic institutions against his express doubts of Christ’s divinity. In the end, a priest did visit, and asked Wilde if he wished to be received; Wilde, who had lost the ability to speak by then, held up his hand, after which the priest baptised him.40 Whether this qualifies as a genuine conversion remains necessarily opaque. It is difficult to imagine a more fitting conclusion for a life of religious ambivalence. The ‘wanderer’ metaphor finds expression in Wilde’s novel, The Picture of Dorian Gray. In chapter 11, the reader learns that the title character is attracted to Catholicism, but never fully convinced: he never fell into the error of arresting his intellectual development by any formal acceptance of a creed or system, or of mistaking, for a house in which to live, an inn that is but suitable for the sojourn of a night in which there are no stars and the moon is in travail.

The religious journey is one of wandering in a starless, moonless night.41 We have been characterizing agnosticism by drawing on metaphors: the suspicious detective, the unmoored sailor, the night wanderer. These illustrate the active ambivalence appropriate to agnosticism, but have negative connotations. Wilde supplies a metaphor with arguably more positive connotation, again in The Picture of Dorian Gray. The character Lord Wotton (chapter 8) says that 38 40 41

39 Quintus (1991), p. 516. Ellmann (1988), pp. 94–5. Ellmann (1988), pp. 583–4. Compare with the verses of Arnold (1851), another agnostic: And we are here as on a darkling plain Swept with confused alarms of struggle and flight, Where ignorant armies clash by night. (Again, my thanks to Mollie Smith for this reference.)

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

98

Agnosticism

religion’s ‘mysteries have all the charm of a flirtation’.42 This is perhaps the best metaphor for Wilde’s ambivalence.43 Reflecting a little upon the concept of flirtation confirms its aptness. Flirtation actively entertains a possibility but never fully accepts it. Sustained acceptance marks the end of flirtation and the beginning of something else. Flirtation by definition is not indifferent, but interested. These descriptors fit quite nicely the interpretation of agnosticism offered so far. As Chesterton puts it, When the Professor is told by the Polynesian that once there was nothing except a great feathered serpent, unless the learned man feels a thrill and a half temptation to wish it were true, he is no judge of such things at all.44

(c) The Place of Belief The above arguments suggest an openness to religion that may include entertaining religious belief or practice episodically, but not accepting a theistic position as rationally convincing or even as overarchingly supra-rational. It appears paradoxical that the rational attitude could include occasional amenability to a position that is (according to the premise of this chapter) overall unconvincing. This has some precedent, as can be shown by an analogy from sport. Anyone who plays baseball knows the feeling of being sure that one is about to hit a pitch. It is a consuming and irresistible feeling of expectancy. It is also usually, or at least often, in error. Even the best hitters only make very solid contact with one or two swung-on pitches of perhaps many more per at-bat. We can list three relevant epistemic features concerning this scenario:

42 Both this quotation from The Picture of Dorian Gray and the in the paragraph before are quoted in Ellmann (1988), p. 94. Note that Wilde’s reference to flirtation makes use of the concept of ‘mystery’, discussed in section III(a). 43 Theism employs similar metaphors: The Old Testament’s Song of Solomon depicts the relationship between God and Israel as a romantic one; the New Testament’s gospels depict Christ as bridegroom, implying that the church is bride. Susanne Langer distinguishes Russell from William James by saying that his essay ‘A Free Man’s Worship’ is ‘never spoiled by any flirtation with the old gods’ (Langer (1979 [1942]), p. 276). 44 Chesterton (1993 [1925]), p. 101.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

99

(1a) Even when correct, the momentary belief is not rationally justified. The batter has too imperfect an ability to predict his or her own movements, let alone those of the ball. (2a) It is also not supra-rational to believe that many swings will result in good contact, because they won’t, and to hold steadily such a position yields no strategic advantage. (3a) Still, there is no need or value in actively discouraging this episodic feeling of being sure of a hit. When entertained episodically, it is perhaps true and not counter-productive if false. (At least, if one assumes a batter of reasonable skill and experience.) Similarly: (1b) A momentary feeling of being sure about God’s existence— sensus divinitatis, for example as Wordsworth describes—is not rationally justified on the agnostic account. (2b) An overall life of faith is not justified on supra-rational grounds, because it doesn’t increase the likelihood of God’s existence, may not be necessary for contact with God, and may interfere with other projects. (3b) Still, there is no need for actively discouraging an openness to the existence of God that may well include occasional feelings of conviction about the divine. This openness may— depending of course on the individual—take the shape of entertaining the existence of God through prayer or ritual of various kinds. Given the agnostic position, this may constitute access to an important truth. Even if such episodic entertainments of faith do not access the truth, given certain parameters, they will not conflict with other values. (Unlike at least one interpretation of Gutting’s reworking of Pascal’s wager, this openness does not require any sustained serious commitment.) Section II in part argued against the central claims of James’ ‘The Will to Believe’.45 This included noting the rigid dichotomy that he and Pascal both rely on in motivating theism. Obviously, the arguments for ambivalence reject that dichotomy. It may be useful to contrast

45

James (2009 [1912]).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

100

Agnosticism

the analogy offered just above with one that James provides in that essay. James claims that belief in God should not follow a model of ‘believing that so-and-so is the case’ but rather should be modelled on the trust we place in people, upon which friendship depends. James argues that the objective evidence available at the outset of a relationship is insufficient for trust.46 Nevertheless, trust is necessary for developing friendships, a condition for the existence of society at large. Our current line of reasoning suggests that we not follow James in his analogy. His argument overlaps somewhat, but not in the right way, with the three points sketched above: As in (1a), so with James: momentary belief in hitting the ball is not justified, nor is an instance of trusting a stranger. In (2a), we claimed that it is not supra-rational to believe that many swings will result in good contact. James, on the other hand, does claim supra-rational justification for an overall attitude of trust, because of its necessity in facilitating relations based on that trust. This would only work as an analogy for religious faith if trusting in God’s existence is necessary for connecting with God. But this chapter has only found good arguments for openness to God’s existence, not for the stronger position of trust in God’s existence. While open-mindedness may be a requirement, an overall attitude of faith is not necessary for contact with God. So, by the lights of this chapter’s arguments at least, James’ comparison to trust and friendship is a faulty analogy. James’ claim would include that there is no need to discourage trust in people, which bears a superficial resemblance to our claims in (3a) that there is no reason to discourage episodic confidence in getting a hit, or analogously, an episodic feeling of conviction about divine existence (3b). But where—on the arguments of this chapter—such belief is merely benign or potentially beneficial, for James the claim amounts to more than this: belief in God’s existence, more than being episodically benign, is a strategic necessity for divine contact. This supra-rational case for belief goes beyond the claims of this chapter, and moreover in a way that is not supported; the baseball analogy and the friendship analogy are thus themselves not analogous to one another.

46

James (2009) [1912]), p. 24.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

101

Now, one might object that since the agnostic position rules out a systemic life of faith it does indeed preclude or at least hinder contact with God. This objection might be mitigated by examining more closely the attitudes/practices consistent with agnosticism.

IV. CONCLUSION: FROM ATTITUDE TO PRACTICE—THE SCOPE OF AGNOSTICISM This chapter has addressed the question of what attitudes are appropriate to agnosticism, in section II by distinguishing it from theism and atheism and articulating its positive philosophical commitments, and in section III via metaphor—investigation, wandering, and flirtation—and analogy, illustrating the rational grounding of openness to particular experiences of God. What actions, or even lifestyles, make sense given the spectrum of agnostic attitudes? Section III(b) covering Oscar Wilde claimed that his ‘protean religious position illustrates the range of attitudes and behaviours available under agnosticism’. This illustration should be accompanied by an account of the principles underlying its aptness. Section III(c)—‘The Place of Belief ’—claimed that the openness appropriate to agnosticism ‘may—depending of course on the individual—take the shape of episodically entertaining the existence of God through prayer or ritual of various kinds’. To what extent? What about individuals otherwise inclined? Section III also raised the possibility that denying the rational justification of sustained belief in God might preclude or hinder contact with the divine. We need to know how inclusive agnosticism is in order to answer this charge. So, now it becomes important to specify—rather than merely illustrate or gesture towards—the scope of rational practices. One limit is distinguished from atheism in only the barest way. This is perhaps what people generally associate with agnosticism, arguably exemplified by Bertrand Russell. As any agnostic must, this person both takes it that there is neither convincing rational nor convincing supra-rational reason to believe in God as a general position, and remains open to entertaining a belief in the existence of God in personal experiences or in rational discourse. Still, this person never does so entertain it; that is, he or she never has such particular experiences, and never finds arguments for God’s existence

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

102

Agnosticism

plausibly compelling. This person prefers not to attend religious rituals or ceremonies, or does so for extra-religious reasons, does not pray, does not reflect on whether God exists, nor discuss it except fleetingly. The other limit, more surprisingly, is distinguished from theism in only the barest way. Again, as any agnostic must, this person both takes it that there is neither convincing rational nor convincing supra-rational reason to believe in God as a general position, and remains open to entertaining a belief in the existence of God in personal experiences or in rational discourse. Moreover, this person does find him- or herself compelled to interpret certain particular experiences as of God, and/or in rational deliberation or discourse entertains theistic arguments as plausibly compelling. This does not result in an overall belief in theism’s truth, and so still counts as agnosticism. Still, this limit case could take the form of a life of regular religious ritual, practice, and reflection, without thereby becoming something other than agnosticism. Many people who describe themselves as religious or even as having faith will fit into this category. Broadly, an overall acceptance of theism, as against agnosticism, has two possible sources, as noted in section II. First, personal experience may compel sustained belief in God. Such experience will not ‘transfer’—it will not compel belief for someone who has not had that experience. In other words, even if it makes sense to the individual in question, it is not rationally demonstrable. Second, one might ultimately accept an argument for believing in God. The agnostic position (as construed in this chapter) does not rule out the possibility that a sound theistic argument exists or could exist, it just reflects that one has not come across such an argument. Of course, as soon as one accepts an argument for a sustained belief in God, whether sound or not, one ceases to be an agnostic. But consider a possible third kind of theist, who bears comparison to an agnostic: They read the critics but on careful reflection do not find them compelling; likewise, though they are aware of theistic arguments and find some of them not without value, they do not believe on the basis of them. Rather . . . it seems to them that they sometimes catch a glimpse of the overwhelming beauty and loveliness of God . . . . After long, hard, conscientious reflection, they find this enormously more convincing

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

103

than the complaints of the critics. Are they then going contrary to duty in believing as they do? Are they being irresponsible? Clearly not . . . They could be wrong, desperately wrong, pitiably wrong in thinking these things. But they are not flouting any discernible duty; they are doing their level best to fulfil their epistemic responsibility.47

Some theists base their faith on being convinced by rational argument that God exists, and others on experience that compels belief. Here the belief is not based on finding a rational argument fully convincing, nor on being compelled by experience, but rather on it seeming that one catches a glimpse of God, and finding this more convincing than critics’ complaints. The experience alone does not compel belief, and the rational reflection upon it does not fully convince, so this form of theism reduces to neither of the other two. Those who pray and engage in religious ritual often will claim to have neither a direct experience of God, nor a conclusive rational demonstration of either God’s existence or the necessity to believe in God. They may describe themselves as theists practicing faith, and this complies with ordinary usage. On the terminology of this chapter, they count as agnostics. They may identify as theists, but the underlying principles are agnostic. Terminology is not the primary issue here. Perhaps the terms need not be exclusive. On this line, one could be an agnostic and a practicing theist. The more fundamental point is that, given the rationality of agnosticism (which we have taken as a premise), it follows that many people who identify as theists have rationally justified attitudes and practices.

BI B LI OGR APHY Arkins, Brian. ‘Mahaffy: Classicist and Philhellen’ (IIHSA Inaugural Address in NUI Dublin, 8 October 1999). http://www.iihsa.ie/mahaffy.html Arnold, M. ‘Dover Beach’. Poetry Foundation (1851) www.poetryfoundation. org (retrieved 21 December 2015). Brown, Frank Burch. The Evolution of Darwin’s Religious Views (Macon, GA: Mercer University Press, NABPR Special Studies Series, No. 10, 1986).

47

Plantinga (1998), p. 214; quoted in Cottingham (2005), p. 125.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

104

Agnosticism

Brown, R.E. ‘Mystery (in the Bible)’. In New Catholic Encyclopedia, edited by Berard Marthaler (New York: Catholic University Press, 2003). Buckley, Michael. At the Origins of Modern Atheism (New Haven: Yale University Press, 1987). Chesterton, G.K. Everlasting Man (San Francisco: Ignatius Press, 1993 [1925]). Clifford, W.K. ‘The Ethics of Belief ’ (1886 [1877]. http://people.brandeis. edu/~teuber/Clifford_ethics.pdf (retrieved 24 December 2015). Cottingham, John. The Spiritual Dimension: Religion, Philosophy, and Human Value (Cambridge: Cambridge University Press, 2005). Cupitt, Don. ‘Interview’. In Conversations on religion, edited by Mick Gordon and Chris Wilkinson (New York: Continuum, 2008). Dulles, A. ‘Mystery (in Theology)’. In New Catholic Encyclopedia, edited by Berard Marthaler (New York: Catholic University Press, 2003). Ellmann, Richard. Oscar Wilde (New York: Alfred A. Knopf, 1988). Garvey, Brian. ‘Absence of Evidence, Evidence of Absence, and the Atheist’s Teapot’. Ars Disputandi 10, no. 1 (2010), pp. 9–22. Grayling, Anthony Clifford. ‘Interview’. In Conversations on religion, edited by Mick Gordon and Chris Wilkinson. (New York: Continuum, 2008). Gutting, Gary. Religious Belief and Religious Skepticism (London: University of Notre Dame Press, 1982). Gutting, Gary. ‘Pascal’s Wager 2.0’. New York Times, The Stone (28 September 2015). http://opinionator.blogs.nytimes.com/?s=pascal%27s+wager +2.0 (retrieved 17 February 2016). Holland, Merlin. Oscar Wilde: A Life in Letters (New York: Carroll and Graf, 2006). James, W. ‘The Will to Believe’. In The Will to Believe and Other Essays in Popular Philosophy (Project Gutenberg, 2009 [1912]). https://www. gutenberg.org/ebooks/26659 (retrieved 21 December 2015). John Paul II, Pope. Encyclical Letter, Fides Et Ratio, of the Supreme Pontiff John Paul II: To the Bishops of the Catholic Church On the Relationship between Faith and Reason (Washington, DC: United States Catholic Conference, 1998). Kors, Alan Charles. Atheism in France, 1650–1729: Volume 1: The Orthodox Sources of Disbelief (Princeton: Princeton University Press, 1990). Lane, Christopher. The Age of Doubt: Tracing the Roots of Our Religious Uncertainty (New Haven: Yale University Press, 2011). Langer, Susanne. Philosophy in a New Key: A Study in the Symbolism of Reason, Rite, and Art (Cambridge, MA: Harvard University Press, 1979 [1942]). Melville, Herbert. Moby Dick (London: Oxford University Press, 1998 [1851]). OED. The Compact Edition of the Oxford English Dictionary (Oxford: Oxford University Press, 1971 [1889]).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

If Agnosticism, Then What?

105

Pascal, Blaise. Pensees (231–241). In Introduction to Philosophy: Classical and Contemporary Readings, edited by John Perry, Michael Bratman, and John Martin Fischer, tr. W.F. Trotter (New York: Oxford University Press, 2016). Plantinga, Alvin. ‘Religion and Epistemology’. In Routledge Encyclopedia of Philosophy, vol. VIII, edited by E. Craig (London: Routledge, 1998). Plantinga, Alvin. Warranted Christian Belief (New York: Oxford University Press, 2000). Plantinga, Alvin. ‘The Reformed Objection to Natural Theology’. In Philosophy of Religion: Selected Readings (3rd ed.). Plenary Session—Philosophical Knowledge and Theological Knowledge, edited by M. Peterson, W. Hasker, B. Reichenbach, and D. Basinger (London: Oxford University Press, 2006 [1980]), pp. 261–73. Quintus, John Allen. ‘Christ, Christianity, and Oscar Wilde’. Texas Studies in Literature and Language 33, no. 4 (1991), pp. 514–27. Russell, Bertrand, Barry Feinberg, and Ronald Kasrils. Dear Bertrand Russell . . . a Selection of His Correspondence with the General Public, 1950–1968 (Boston, MA: Houghton Mifflin, 1969). Smith II, Philip E., and Michael S. Helfand. Oscar Wilde’s Oxford Notebooks: A Portrait of Mind in the Making (New York: Oxford University Press, 1989). Wilde, Oscar. The Picture of Dorian Gray (New York: Tess Press: N.d. [1891]). Wolf, Gary. ‘The Church of the Non-believers’. Wired. November 2006. https:// www.wired.com/2006/11/atheism/?pg=1&topic=atheism&topic_set= Wootton, David. The Invention of Science: A New History of the Scientific Revolution (New York: Harper-Collins, 2015). Wordsworth, William. ‘Lines Composed a Few Miles above Tintern Abbey, on Revisiting the Banks of the Wye during a Tour, July 13, 1798’ (1798). Academy of American Poets. www.poets.org (retrieved 21 December 2015).

F U R T H E R RE A D I N G Critchley, Simon. ‘Oscar Wilde’s Faithless Christianity. The Guardian (15 January 2009). www.theguardian.com Gray, John. ‘Interview’. In Conversations on religion, edited by Mick Gordon and Chris Wilkinson. (New York: Continuum, 2008). Ludin, Roger. Emily Dickinson and the Art of Belief (Cambridge: William B. Erdmans, 1998).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

106

Agnosticism

Mougin, Gregory, and Elliot Sober. ‘Betting against Pascal’s Wager’. Nous 28, no. 3 (1994), pp. 382–95. Wilde, Oscar. The Complete Works (London: Barnes and Noble: 2007). Woods, Patrick. ‘From the Middle Out: A Case for Agnosticism’. Sophia 46 (2007), pp. 35–48.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

4 ‘New Agnosticism’, Imaginative Challenge, and Religious Experience David Leech

I. INTRODUCTION This chapter makes a contribution to recent discussions within what might be called new conceptual territory for agnosticism opened up by John Schellenberg’s pioneering work in the philosophy of religion. Schellenberg has argued that neither belief nor disbelief in any detailed religious proposition is justified at this relatively very early stage in religious enquiry, but non-doxastic imaginative faith in what he terms ‘ultimism’ is ‘peculiarly appropriate’ for humans who may yet have vast periods of future religious investigation still ahead of them.1 By ‘ultimism’, Schellenberg refers to the religious proposition that there is a reality which is metaphysically, axiologically, and soteriologically ultimate. In other words, should it exist, it would have ontological fundamentality, unsurpassable objective value, and would be that in relation to which humans achieve their highest good.2 By ‘non-doxastic faith’ he means having faith that x, without being able to go so far as believing x, where one would be within one’s 1

See Schellenberg (2013a), p. 144. Ultimacy implies transcendence. The precise definition of ‘transcendent reality’ Schellenberg offers is: ‘something . . . more than or other than—if it in some ways goes beyond—the physical universe studied by science. (It need not oppose physical nature; it might indeed include it in some sense, if that is conceivable.) . . . [it] will not be ontologically or causally reducible to the universe.’ (2005), p. 26. 2

David Leech, ‘New Agnosticism’, Imaginative Challenge, and Religious Experience In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0005

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

108

Agnosticism

epistemic rights to have faith that x, but not in believing (or disbelieving) x. For instance, a golfer may have faith that she will hit a perfect shot, choosing to push away a certain amount of doubt— which, however, doesn’t amount to justified disbelief in her ability— to imagine her perfect swing, the flight of the ball and its perfect landing, without being within her epistemic rights in believing she will achieve such a thing. Similarly, one could have faith that such and such a religious picture—here, ultimism—is the case, while lacking the kind of evidence that would justify belief or disbelief in it. Schellenberg has also argued that such non-doxastic faith in ultimism can ground a robust imagination-based way of being religious in which humans, motivated by this non-doxastic faith, can ‘strain . . . to the limits’ and achieve ‘intrinsically valuable states of mind’.3 Indeed he proposes that ultimism—which, he argues, also entails that there must be an afterlife—encourages an attentiveness to evidence for goodness and deeper significance in reality, which would combine contemplative and reverential elements.4 It also (he supposes) justifies efforts to diminish anxiety about finite personal goods and promote risk-taking behaviour for the sake of the good.5 (Nor is it incompatible with participation in traditional religious faith communities which also encourage these goals, provided of course that the ultimist’s nondoxastic faith is not made a ground for exclusion.) However, there has been some recent debate about whether, granted the appropriateness of suspending belief and disbelief towards ultimism at this early point in human religious enquiry, Schellenberg has nevertheless been too quick to suppose that ultimism is the most adequate non-doxastic faith stance for a person who is agnostic in this sense, or that it is the most adequate proposal for a new research programme in the philosophy of religion, since there are other options within what can be called a broad ‘religious pluralism’ tradition both for non-doxastic faith and for religious enquiry.6 Without making any attempt to be exhaustive, I will consider in what follows two possible challenges to ultimism as a non-doxastic 3 Schellenberg (2009), p. 115. For an introduction to Schellenberg’s approach in the philosophy of religion, see for instance the summary by Draper (2011), pp. 59–65, and Schellenberg (2013a). 4 5 Schellenberg (2009), pp. 37–8. Schellenberg (2009), pp. 40–1. 6 See Diller (2013), esp. p. 229: ‘Sceptical ultimism is a new member of a secondorder class of religious forms that incorporate other first-order religious forms. It joins venerable members of this class such as Bahá’í and Unitarianism.’

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

109

faith stance and as a research programme from within this ‘tradition’: a potential challenge from (1) Kantian-inspired religious pluralism; and (2) from the claim that some broader religious framework proposition (for instance, the more minimalist claim that ‘there is something (religiously speaking)’) is more adequate for these purposes than ultimism.7 I conclude that notwithstanding the prima facie force of these two objections (or at least, of objection (2)), it is still appropriate for an agnostic (or, more properly, a former local agnostic in a sense to be defined immediately below) to embrace ultimistic non-doxastic faith.

II. SCHELLENBERG’S ‘NEW AGNOSTICISM’: ULTIMISTIC NON-DOXASTIC FAITH In a reply to Peter Forrest concerning a reference of Forrest’s to New Atheism, Schellenberg writes that the adjective ‘new’ would be better applied to agnosticism about ultimism, since ‘[t]he old agnosticism concerned only God. The new agnosticism, which can absorb ordinary or old atheism, and so can reject the old agnosticism, kicks things up a notch, focusing broadly on ultimism rather than narrowly on theism.’ By ‘old agnosticism’, he means traditional agnosticism, that is, the assumption of an agnostic stance towards the question of whether or not theism is true (we might also call this ‘local’ agnosticism). However, by ‘new agnosticism’, he means agnosticism in the sense of suspension of the propositional attitude of belief or disbelief about his own novel proposal in the philosophy of religion, ‘ultimism’.8 This is the sense in which Schellenberg himself is an agnostic—he neither believes nor disbelieves in ultimism because he considers that the state of the evidence, at this stage in philosophical enquiry into religion, does not permit him to believe either that it is true or that it is false. Ultimism is a broad religious framework proposition, but it can be filled out or ‘elaborated’ in various ways. By ‘elaborated’, Schellenberg means that ultimism can be given detailed content—for instance, as the Trinitarian God of Christians, the Dao, Buddha Nature, etc.—or To my knowledge the first has not yet been formulated; versions of the second have recently been formulated by Elliott (2017), pp. 97–116; and, implicitly, by Diller (2013). 8 Schellenberg (2013b), p. 695. 7

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

110

Agnosticism

in numerous as yet unanticipated ways in the future. However, a Schellenbergian ultimist must refrain from assenting to any elaborated form of ultimism, having non-doxastic faith solely in the most generic expression of it, since humans are not yet at the point where they could choose with any rational justification between elaborated ultimisms, and having faith in the most generic form maximizes the chances of being broadly correct at this early stage of human religious investigation. By ‘kick[ing] things up a notch’ Schellenberg means having an agnostic attitude towards belief or disbelief in this broad religious framework proposition, rather than just having an agnostic attitude towards belief or disbelief in theism. Also, because Schellenberg thinks that one elaborated version of ultimism—classical theism—can already be ruled out at this stage of philosophical enquiry into religion, he thinks that this new agnosticism can ‘absorb’ old (or ‘local’) atheism, and so leave behind the ‘old’ agnosticism which only considered whether theism or local atheism was the case.9 Schellenberg nevertheless believes that although belief in ultimism is not justified at this stage in philosophical enquiry into religion, ultimistic non-doxastic faith is justified. In other words, he does not think that the erstwhile local agnostic needs to sit tight with pure agnosticism (what he terms ‘pure scepticism’) without a religious faith commitment of any sort.10 There are several reasons why Schellenberg does not suppose that pure scepticism is preferable to ultimism. First, the pure sceptic will miss out on a range of non-truthoriented factors—the intrinsically valuable dispositions and states of mind mentioned above—which favour non-doxastic ultimistic faith. Second, by adopting such faith, the ultimist nevertheless keeps ‘one eye open’ for arguments which may reveal the falsehood of ultimism, so that epistemically she remains on a par with the pure sceptic with respect to her ability to track metaphysical truth rationally. In other words, non-truth-oriented factors are not trumping truth-oriented ones here.11 Third, not being faithful to one’s tendentially ultimistic

9 Schellenberg notes that ‘not all advocates of evolutionary religion will be atheists, denying traditional theism; some will be agnostics instead’ (2013c), p. 134. I here use the term ‘local atheism’ (and by extension, ‘local agnosticism’) in Draper’s (2017) sense. 10 ‘a purely skeptical response to ultimism is a response involving passive and/or active categorical religious skepticism but without the admixture of faith’ Schellenberg (2009), p. 4. 11 Schellenberg (2009), p. 84.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

111

hunches/experiences—if one has them—and imaginatively developing them would reveal a distrust in one’s positive responses to the world.12 Schellenberg supposes that we can move beyond pure scepticism because ultimism also has ‘the good of being true to reason’, since ‘[r]easoning indicates that ultimism is just as likely to be false as true, or that the relevant likelihoods cannot be determined.’13 In short, we can voluntarily adopt ultimistic non-doxastic faith and live as though it were true, even though the state of the evidence does not permit us to believe (or disbelieve) it.14 He also suggests that this ‘new agnosticism’ concerning ultimism could form a, or the, major research programme in the philosophy of religion, superceding the narrower traditional debate between atheism and theism. Schellenberg provides two principal reasons why ultimism should be taken both as the object of philosophical religious enquiry and as an object of religious beliefless faith, namely, that a religious framework proposition should (a) have some explanation for the limittranscending strong religious experiences testified to in all religious traditions; and (b) should ‘challenge’ us, that is, make us pursue inprinciple infinite goals (I will return to these below). In this sense Schellenberg has done much to develop the agnostic position in the philosophy of religion, even if the term ‘agnostic’ is not often used by him (he prefers ‘religious sceptic’), presumably because in its normal sense, the term, like ‘atheist’, has local connotations which he would prefer to avoid, that is, that it is an attitude towards the traditional theism/atheism debate narrowly conceived. But with these caveats, we can say that Schellenberg’s work makes a major contribution to what have traditionally been designated as ‘agnostic’ stances to the philosophy of religion. A distinctive feature of Schellenberg’s position is his emphasis on humans’ place in time, and its implications for religious enquiry. If, as

12 See section IV below for a more extended discussion of this third reason. Of course, for the person who does not have such hunches/experiences—even though Schellenberg claims that they are very widespread—only the first two reasons would count as factors favouring her adoption of ultimistic non-doxastic faith rather than pure scepticism. 13 Schellenberg (2009), p. 76. 14 Cf. Robin Le Poidevin’s observation that the agnostic can have a religious commitment, since even fictions can engage emotions and action, and indeed the agnostic is in the situation of not knowing whether the religious ‘stories’ that she could choose to entertain might correspond in whole or part to reality. Le Poidevin (2010), pp. 105–6.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

112

Agnosticism

he notes, the 200,000 or so year history of homo sapiens—of which only around 6,000 years has seen systematic religious enquiry—is wedged between three and a half billion years of evolutionary development on one side—life’s past—and another billion on the other—life’s potential future . . . What developments in religiously-relevant thought and feeling might Earth see in so much time?15

Schellenberg argues that transitioning to seeing human religious enquiry within these deep time perspectives should lead us to adopt only general religious ideas (such as ultimism), since any detailed religious proposals are likely to be reversed by future developments especially as they are informed by more of the total evidence.16 On the face of it this transition to carrying out religious enquiry mindful of these deep time perspectives would have the following implications for local agnosticism: the agnostic who believes the current evidence for religion/non-religion is equally likely will need to be sceptical about whether the total evidence will be similarly equiprobable. And the agnostic who believes that the relevant evidence is largely absent or even in-principle out of reach, should be prepared to revise this belief in light of the possibility that over the stupendously long period of potentially another billion years of enquiry it may emerge that there are good reasons for revising how out of reach the evidence is, or doubting its in-principle inaccessibility.17

III. A POTENTIAL CHALLENGE TO ULTIMISM FROM KANTIAN-INSPIRED RELIGIOUS PLURALISM Suppose a local agnostic gradually relinquishes her local agnosticism and begins to consider a range of broad religious positions towards 15

Schellenberg (2013c), p. 3. By total evidence, Schellenberg understands all the evidence, recognized and unrecognized, which is relevant to the justification of some belief x. Temporalism gives contemporary human beings excellent reasons to worry whether they have access to the total body of relevant evidence relative to x, since they are a mere 6,000 years or so into religious enquiry which could potentially extend another billion years into the future. See for instance Schellenberg (2007), pp. 28–9. 17 For instance Draper notes that an agnostic can judge that religious evidence is ambiguous on the distinct grounds that it is ‘largely absent’, ‘present but vague’, or equiprobable (Draper (2002), pp. 197–8). However, all of these judgments look fragile and subject to revision in a temporalist perspective. 16

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

113

which she can adopt non-doxastic faith within what Diller has referred to as a second order set of religious forms which ‘handle multiple religious traditions with even-handed respect in a way that many of us want.’18 How would she orient herself?19 Consider first her possible attitude to classical religious pluralism as represented by Hick’s ‘pluralistic hypothesis’. This is not a religious framework proposition in Schellenberg’s sense, since it is too specific. However, it is also not an elaborated ultimism. As Schellenberg himself notes, if Hick’s religious pluralism were correct, then ultimism itself could not literally be true, since ultimism presupposes realism and Hick’s Kantian religious pluralism entails that humans could never discover that some realist religious possibility was the case, since such a discovery would be epistemically impossible for humans.20 Hick may in practice fail to adhere to his own Kantian principles—Insole for instance has argued that he may actually only be committed to the trivial thesis that the mind makes contributions to the human experience of reality, which is compatible with realism, and he also makes substantial claims about the Real21— but for the purposes of this discussion I take Hick to be committed at least officially to an interpretation of Kantian transcendental idealism as comprising two separate realms (noumenal, phenomenal), and to denying that it is epistemically possible for humans to have true substantial (not merely formal) beliefs about the nature of the noumenal. Insofar as he denies the latter, I take this to be an anti-realist aspect of his position.22 In this sense Hick’s position is in competition with ultimism, rather than an (elaborated) instance of it. 18

Diller (2013), p. 229. Henceforth I will generally refer to this hypothetical agnostic simply as ‘agnostic’ rather than ‘former local agnostic’, although the latter should be tacitly understood. 20 ‘[U]ltimism, unlike the concept of the ‘Real’, is not insensitive to real-world aspirations to gain some understanding of God’ (Schellenberg (2016), p. 176). 21 See Insole (2006), pp. 121–2, 133. 22 In this respect I follow Insole, who defines anti-realist approaches to religion as those (and only those) involving the denial of at least one of four criteria, the last of which (‘D’) is that ‘we can, in principle, have true beliefs about what is the case independent of human cognition’ (Insole (2006), p. 2). Insole notes that ‘In as much as Hick does not abandon the noumenal/phenomenal and formal/substantial distinctions, his realism about God is undermined by the in-principle impossibility of arriving at any truths about God that are true independent of human cognition. It is not desirable that the only thing that saves Hick from anti-realism (in that he denies D) is his preparedness to break his own rules and covertly make substantial claims about God’ (Insole (2006), p. 134). 19

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

114

Agnosticism

But we may ask what the prospects are for non-doxastic faith in religious pluralism instead of in ultimism? Hick, like Schellenberg, regards the universe as religiously ambiguous. In assessing the respective evidence for and against a religious hypothesis, Hick of course presents a very similar ‘balance sheet’ account to Schellenberg, concluding similarly that the (present) total evidence does not decisively favour either naturalism or a religious interpretation of reality, so that the universe remains ambiguous.23 And just as Schellenberg adds ultimistic non-doxastic faith to pure scepticism, so Hick adds a second-order religious proposition, the Kantian Real, to his scepticism, passing beyond agnosticism to belief in a religious interpretation of reality. Hick differs from Schellenberg here not only in positing the ‘Real’ (noumenal) instead of generic ultimism, but with respect to his propositional attitude towards it, proposing that belief in the ‘pluralistic hypothesis’—that is, that the world religious traditions are equally valid expressions of the Real—is rational at this stage in human religious enquiry. However, this can also quite easily be modified into a non-doxastic commitment to a Kantian religious pluralism. What would motivate such an agnostic to opt for ultimistic nondoxastic faith rather than its Hickean equivalent? Schellenberg has provided some reasons of his own for preferring ultimism over Kantian religious pluralism, and these reasons apply equally well to the case of non-doxastic faith in Kantian religious pluralism, namely, that although it is in his view well-motivated, it also goes ‘too far’, because its extreme ineffability means it is vague to the point of uselessness (what I have designated as the anti-realist aspect of Hick’s position, at least insofar as he officially defends the noumenal/ phenomenal and formal/substantial distinctions). By contrast, ultimism accommodates the fact of religious diversity without putting an in-principle block on future religious investigation: we humans have no obvious way to tell, at our early stage in religious investigation, that religious insights are certainly beyond our epistemic reach.24 One can also add that other forms of religious pluralism can replace Hick’s (official) Kantian religious epistemology while still doing the same explanatory work. For instance, Peter Byrne has developed a more unambiguously realist form of pluralism in order

23

Hick (2005), p. 124.

24

Schellenberg (2016), p. 176.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

115

to obviate what he sees as a number of problems arising specifically from Hick’s Kantian-inspired (if sometimes un-Kantian in practice) ‘pluralism hypothesis’ and the vacillation between realism and antirealism to which it is prone.25 Whatever the merits of Byrne’s position, for the purposes of a temporally contextualized argument of this kind—that is, mindful of deep time perspectives—it is not necessary to outline and defend it in any detail. We may simply acknowledge that it and similar alternative religious pluralist models exist, they have been intelligently defended, are not manifestly selfcontradictory, and that for all we know, they or similar versions of non-Kantian religious pluralism could be convincingly elaborated over the potentially billion years of possible future religious enquiry. So construed, then, the class of explicitly non-Kantian religious pluralisms can also take its place as a class of elaborated ultimisms. Why, in that case, should our agnostic, faced with the choice between them, choose Kantian religious pluralism over unambiguous realism (including the sub-class of realist religious pluralisms)? Any preference for Kantianism in these circumstances can just seem arbitrary, since it is necessary neither for ethico-political reasons (i.e. defending the value of tolerance) nor for explaining the range of putative facts (i.e. salvific parity, a common core of religious experience, etc.) which it is meant to explain. This is because for all we know, some elaborated ultimism belonging to the sub-class of unambiguously realist religious pluralisms may do an equally good job. At this point, however, the defender of a Kantian religious pluralism could perhaps appeal to the need for a consistent agnostic to apply their agnosticism to metaphysical as well as religious belief, reasoning like this: given the impressive historical and contemporary peer disagreements about metaphysical worldview stances (realism, anti-realism, Kantian transcendental idealism, internal realism, etc.) the agnostic should be as agnostic about these stances as she is about religious/non-religious ones. Perhaps such a defender could also argue that Schellenberg’s position is self-defeating, since he fails to apply temporalist scepticism to his own realist assumptions, but if he 25 Byrne, like Hick, argues from the ubiquity of religion, the universality of belief in a sacred reality, and salvific parity to the plausibility of affirming one transcendent religious reality. But unlike Hick, he remains unambiguously realist, doing so by adopting a causal rather than descriptive theory of reference to get around the problems of reconciling pluralism with realism. See Byrne (1995); and discussion in Lehe (2014), pp. 505–20.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

116

Agnosticism

were to do this, he would have to acknowledge that metaphysical beliefs are as unjustified as religious beliefs, so at most he should have only non-doxastic faith in realism. This would not provide a reason for favouring Kantian transcendental idealism over realism as a nondoxastic metaphysical faith stance, but neither (so it could be argued) does it provide a reason for favouring realism (and therefore ultimism as a realist religious framework proposition). However, Schellenberg has a more precise reason for not favouring the Kantian metaphysical option, namely, that it falls, together with religious beliefs, into the class of beliefs characterized by precision, detail, profundity, attractiveness, ambition, and controversiality.26 This (as he notes) makes it particularly subject to temporalist scepticism. Although he does not spell it out, presumably Schellenberg would hold that conversely realism meets fewer of these criteria. It is difficult to know how to apply all of these criteria to realism uncontroversially, but at least it seems plausible to say that realism is less ambitious and controversial than Kantianism, and perhaps less profound (?) too. But Schellenberg could possibly also argue as follows: realism as a metaphysical stance is a natural elaboration of the naïve realist beliefs which humans unavoidably develop, and for this reason it is the sort of thing we humans can reasonably continue to believe, so his commitment to realism is not self-defeating.27 For instance, in Evolutionary Religion Schellenberg defends belief in (apparently) well-established scientific theories on the grounds that some results of serious enquiry, which are both ‘well-evidenced’ and ‘obviousseeming’ may also—like unavoidable natural beliefs—be ‘impossible not to believe’ for careful inquirers. It may therefore be ‘natural and appropriate’ for humans at this point in our evolutionary trajectory to treat such ‘obvious-seeming’ things as worthy of belief.28 Schellenberg has defined this class of potentially belief-worthy claims sufficiently broadly (i.e. as well-evidenced, and obviousseeming, to a careful inquirer) that one can perhaps imagine him 26 Schellenberg (2013c), p. 52. Schellenberg does not mention ‘attractive’ (also in his list) in connection with Kantianism. 27 This is why global scepticism does not threaten: ‘the skeptic is unable to prove that our unavoidable belief-forming processes are unreliable. All things considered, then, we are quite reasonable in regarding those belief-forming practices as reliable for the sake of inquiry’ (Schellenberg (2013c), p. 38). 28 Schellenberg (2013c), p. 41.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

117

treating realism too as just such an obvious-seeming belief-worthy claim, so that it would be exempt from what Forrest has called ‘Schellenberg’s razor’—the restriction of belief to merely unavoidable beliefs.29 Diller also notes that Schellenberg’s evolutionary religion avoids many of the problems with Hick’s religious pluralism. First, ultimism represents a substantial metaphysical position, positing an actual sacred ‘object’ which for all we know may exist and be known at some point in the future, rather than Hick’s (officially) inaccessible Real. Second, while Hick’s critics complain that Hick tends generically to reduce salvation to the movement from self- to realitycenteredness, Schellenberg can be content to say that the various religions have distinct ways of filling out the soteriological as well as the metaphysical and axiological dimensions, one of which (or some future variant) may turn out to be the case after the long progress of religious enquiry stretching into the deep future. Therefore ultimism does not close down religious enquiry in these respects in the way Kantian religious pluralism does.30

IV. THE CHALLENGE TO ULTIMISM FROM A BROADER RELIGIOUS FRAMEWORK PROPOSITION (IETSISM) Suppose that the agnostic is persuaded that ultimism is not selfrefuting for the reasons just considered, and that prioritizing realist framework propositions which do not close down religious enquiry in the way Hick’s pluralism does is appropriate. However, it may yet be the case that even if there is no compelling reason for her to adopt non-doxastic faith in religious pluralism, she should nevertheless have non-doxastic faith in some other religious framework proposition than ultimism. 29 See Forrest (2013), p. 540. Rottschaefer’s concern is therefore misplaced that the Schellenbergian razor might shave off belief in scientific theories, with the result that temporalism itself (based on cosmological and biological scientific theories) and the idea of ‘total evidence’ would be undermined. See Rottschaefer (2016), p. 493, ft.11. 30 ‘This more general framing lets Schellenberg tolerate more diversity in his species than Hick, and thus better respects the species’ particularity’ (Diller (2013), pp. 229–30).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

118

Agnosticism

Schellenberg makes the following two claims in favour of ultimistic rather than some other religious non-doxastic faith: (1) ultimism is about as general a claim as one could adopt without leaving religion altogether—general enough, indeed, to permit investigation of all other strong religious claims, which represent the various ways of filling it out. (2) evolutionary religion will need to find some place for powerful religious experiences and also remembering that it should challenge us [sc. existentially and imaginatively].31

However, Diller and Elliott have recently argued with respect to (1) that in fact Schellenberg’s ultimism arbitrarily restricts the scope of ‘religious’ to ultimist religious conceptions, and thereby runs the danger of prematurely closing down investigation into religious possibilities (and Rottshaefer, a religious naturalist, has noted that it leaves out forms of religious naturalism).32 In other words, it is not the most general religious claim. I will follow Elliott here in referring to this broader religious framework proposition as ‘ietsism’ (an Anglicized version of ‘ietsisme’, the Dutch for ‘somethingism’). He characterizes ietsism as the nondoxastic faith that ‘there is merely something worth committing ourselves to religiously’ and claims that it can meet Schellenberg’s criteria of imaginative challenge and capacity to explain powerful religious experiences.33 He further argues that it turning out to be the case that there is something worth committing ourselves to religiously is less incredible than some elaborated ultimism turning out to be the case. More precisely, he defines his version of ietsism as follows: ‘There exists a soteriologically transcendent reality that may or may not also be axiologically and/or metaphysically transcendent (and very little more can be said regarding its nature)’.34 He also notes that it includes, but is more extensive than, ultimism, since it will include any internally consistent religious possibilities which exhibit transcendence in any one, any two, or all three of the Schellenbergian

31

Schellenberg (2013c), pp. 99, 103 (numbering added by author). ‘The objects of religious enquiry and practice have not and need not be ontologically transcendent, that is, they might be ontologically natural (that is, using Schellenberg’s definition, possible objects of scientific study)’ (Rottschaeffer (2016), p. 484). 33 34 Elliott (2017), p. 98. Elliott (2017), p. 98. 32

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

119

dimensions (metaphysical, axiological, soteriological). Importantly, he adds that from a prior probability standpoint, it must logically be more likely that some elaborated form of ietsism is the case than that some elaborated form of ultimism is. This is because ietsism simply includes a much greater range of religious possibilities.35 Elliott also claims that Schellenberg is mistaken in thinking that ultimism is peculiarly suited to accounting for powerful religious experiences, since he argues that ietsism can handle the data of religious experience just as well as can ultimism.36 In this he follows Diller, who argues that while Schellenberg acknowledges that there are non-ultimistic forms of religion (e.g. the pantheon of limited gods in extinct Greek and Norse religions, or in forms of living Hindu belief, etc.), he ‘undersells’ their importance because he finds a tendency to ultimize in them.37 Finally, Elliott argues that the imaginative challenge requirement—that a religious framework proposition should challenge human beings to enable the richest possible development of their personalities—can also be just as much (or nearly as much) satisfied by ietsism as by ultimism. But in any case—and this is Elliott’s main point—the logically greater likelihood of ietsism’s being true should trump the imaginative challenge consideration, since we ‘can more robustly be moved by something we find more likely to be true.’38 He also observes that the way is open for the person of ietsistic non-doxastic faith to nevertheless hope that one elaborated ultimism within the subset of ultimist religious possibilities turns out to be the case, since nothing stops an ietsist from hoping for this.39

35 Elliott (2017), pp. 99–100. Elliott also notes (p. 114) that ‘Nearly any religious or quasi-religious position . . . falls under the scope of IETSISM.’ By contrast, Schellenberg insists that it is rather ‘by embracing these three together [sc. metaphysical, axiological, and soteriological dimensions] . . . that we might profitably see religiousness as instantiated’ (Schellenberg (2016), p. 166). 36 Elliott (2017), p. 99. 37 Diller (2013), p. 226. She does not mention religious experience here, but her general point is that Schellenberg divides religious traditions up into ones which actually (‘fully’) ultimize, and ones which potentially or tendentially do so, thereby excluding genuinely non-ultimistic religious forms: ‘The implication is that, all told, religions are ultimizing enough that the field can safely ignore the non-ultimistic elements in them.’ 38 Elliott (2017), p. 106. By contrast he sees Schellenberg as prioritizing the ability of a religious framework proposition to challenge imaginatively (pragmatic considerations) over its prior probability (truth-oriented considerations). 39 Elliott (2017), p. 112.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

120

Agnosticism

What then should the agnostic make of Schellenberg’s exclusion or, more accurately (in his own words) ‘sending to the back of the line’ of non-ultimistic religious conceptions? Given that ietism includes ultimism in addition to other generic religious positions, it may seem sensible to the agnostic who is impressed by temporalism to suppose that with respect to religious enquiry, it would be advisable not to delimit it exclusively to ultimistic possibilities, and with respect to adopting a non-doxastic faith stance, to embrace ietsism and at most to hope that ultimism is the case rather than have non-doxastic faith that it is. However, granted the bare logical point that it is more likely that some elaborated form of ietsism is the case than that some elaborated form of ultimism is, it seems to me that it is nevertheless open to an ultimistic reply in a number of ways. In what follows I will briefly note the pragmatic considerations which an ultimist could appeal to in this connection, but I will then pass on to a longer discussion of the truth-oriented considerations. In terms of Schellenberg’s two abovementioned claims in favour of ultimistic rather than some other religious non-doxastic faith, his contention (in claim 2) that a religious conception ‘should challenge us’ alludes to pragmatic considerations. But his reference (also claim 2) to the need for a religious conception adequate to our early stage in religious enquiry to find ‘some place for powerful religious experiences’, and to constitute ‘as general a claim as one could adopt without leaving religion altogether’ (claim 1)—both true-oriented considerations—are most relevant, since I believe the most interesting objection here is the one that ietsism is finally preferable because humans can more easily have non-doxastic faith in the position they see is more likely to be true. If humans can ‘more robustly be moved’ by something they find more likely to be true, it seems prima facie quite compelling that this should override the imaginative challenge (i.e. pragmatic) factor in importance, since ietsism has a higher initial probably than ultimism. But I suggest that an ultimist does not even need to rely on non-truth-oriented factors to make the case for ultimism’s preferability to ietsism. I therefore consider first the internal consistency of ietsism as a position; then how it fares relative to ultimism in explaining religious experience. Finally I consider what I suggest is the most difficult challenge to ietsism, the question of its viability as a faith stance. I argue that ultimism rather than ietsism may in the end be ‘as general a claim as one could adopt’ (in the sense

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

121

of adopting a faith stance, rather than as a programme for religious enquiry), not because one would otherwise ‘leave religion altogether’, but because ietsism may be unviable as a non-doxastic faith stance and a religious guide to behaviour. One non-truth-oriented reason an ultimist can appeal to for preferring ultimistic non-doxastic faith to the ietsistic equivalent is its greater potential to generate religious zeal (and in this sense ‘challenge us’). Elliott’s claim for instance that the difference in the capability of the two faith stances to generate religious zeal may be minimal seems questionable.40 Schellenberg’s appeal (in Jamesian vein) to an ought-to-be-true argument for ultimism, that is, that ‘it is appropriate for us to desire that [ultimism] be so’ seems to work better for ultimism than ietsism.41 Suppose at some distant future time x a consensus is reached that some elaborated non-ultimistic religious possibility is actually the case. It is easy to imagine a person who had formerly lived according to an ultimistic non-doxastic faith experiencing joy that metaphysical naturalism is not the case42 but with an admixture of sadness on discovering that something limited like, for example, Aristotle’s Unmoved Mover (Elliott’s own example) is the case.43 (The limited religious reality in question may also fail to give humans a reason to suppose that, for example, all human suffering and tragic foreshortening of life will be atoned for in the end, on account of some form of post-mortem survival, etc.) Schellenberg also proposes the following additional pragmatic consideration favouring ultimism, namely, that there is ‘something noble in an imagination-based life of the sort in question [sc. lived with ultimistic non-doxastic faith], in which a finite individual is found always straining to the limits’, and which allows for the fullest development of the whole person, bringing to human life an ‘infinite’ quality.44 But it would seem that bringing to human life an ‘infinite’ quality is something which only an ultimistic non-doxastic faith can

40

41 Elliott (2017), p. 107. Schellenberg (2009), p. 228. On the assumption that naturalism is objectively saddening. 43 Of course the ietsist will observe that her non-doxastic faith is not in some nonultimistic religious possibility turning out to be the case, but in some religious possibility simpliciter, whether ultimistic or non-ultimistic. But I will return to difficulties with this response in my final point about the adequacy of ietsistic nondoxastic faith. 44 Schellenberg (2009), pp. 115, 175. 42

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

122

Agnosticism

achieve.45 The case of ietsism is admittedly complex, since it also includes ultimism as one subset within its set, although I will suggest in my final point that the ‘unfocused’ nature of ietsism as a nondoxastic faith stance also results in it being unviable as a religious guide to behaviour. But for now I will pass to a consideration of the truth-oriented considerations an ultimist could appeal to which would favour ultimism over ietsism. First, the ultimist could query whether ietsism is internally consistent. One of Schellenberg’s reasons for proposing that only nondoxastic faith in ultimism is justified is that ultimism itself may possibly be inconsistent, in the sense that it is epistemically possible that understanding of the Ultimate is beyond human reach, but the soteriological dimension of the concept of ultimacy implies that the greatest human good is obtained through conscious relation to the Ultimate.46 Showing that for all we know ietsism may also possibly be internally inconsistent would put ietsism on a par with ultimism rather than favour ultimism. However, it would blunt arguments that ultimism was somehow more incredible than ietsism. To return then to Elliott’s precise definition of ietsism: at first sight it is not clear that the position is internally consistent, since Elliott claims that soteriological transcendence (ST) does not entail metaphysical and axiological transcendence (MT, AT)—his model is strictly agnostic about whether the object in relation to which humans may achieve ST is AT and/or MT.47 But it is unclear whether it makes sense to speak of ST without supposing transcendence in one or both of the other two dimensions, MT and AT. Elliott himself maintains that he can ‘see no reason’ why ‘a soteriologically transcendent reality would also, ipso facto, be both MT and AT’,48 but arguably ST requires (at least) MT, and perhaps also AT. What would an example of an elaborated version of ietsism having ST alone, in a set consisting Cf. also: ‘the notion of a limited person—or indeed of a limited anything—does much less to challenge our imagination and encourage the stretching of our conceptual capacities than a strong idea [like ultimism] might do’ (Schellenberg (2013c), p. 98). 46 Schellenberg (2007), pp. 64–8. 47 Elliott (2017), p. 115, ft.31: ‘This seems to be logically possible, to me—in the least, an argument would need to be made that something must be AT and MT if it is to be ST.’ 48 Elliott (2017), p. 107. 45

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

123

of ST, non-MT, and/or non-AT, look like? In terms first of MT: prima facie it is hard to make sense of what soteriological transcendence would look like in a purely ontologically natural world. Elliott invites us to ‘[i]magine that a religious reality is indeed ST: being rightly related to it will make for a greater well-being that what can otherwise “naturally be attained”.’ He then invites us to imagine that it is not MT or AT, and concludes: [w]hy assume that such won’t garner a robust religious zeal? It seems that what’s fundamentally pertinent to garnering existential zeal in religious assent is ST; the object of [evolutionary religion] needs to pose a transcendent value for us, as opposed to constituting some transcendent value or metaphysical category simpliciter.49

Whether this works turns, I suggest, on whether it is consistent to talk of humans achieving a transcendent good-for-themselves by being rightly related to a purely natural object or state of affairs with a purely natural intrinsic value. Elliott is aware of the problem, but thinks it leaves his argument intact: [s]ince Schellenberg defines MT, AT, and ST in terms of nonnaturalness, it may be odd to assert that something can be ST without being MT as well. Likewise, it may seem odd that something ST may not be AT. This is a fair suggestion, and of no consequence for my argument. My main interest is in rejecting the triple-transcendence thesis, which simply denies that MT, AT, and ST are conjunctively required for garnering a religious zeal.

However, I am not sure how this can be of no consequence, since the consistency of ietsism (as Elliott defines it) will enter into a person’s calculation of its epistemic ‘tenability’, and hence—by his own reasoning—affect the religious zeal it can generate. The difficulty, I suggest, cannot be resolved simply by rejecting Schellenberg’s terms, but turns on whether talking of ST in relation to a natural object/state of affairs is consistent.50 Many will find it intuitive to suppose that a person would only achieve a transcendent good by being rightly related to a transcendent object/state of affairs, and 49

Elliott (2017), p. 108. Elliott (2017), p. 115, ft.30. Possibly, however, Elliott means to use ‘transcendence’ in a sense which is sufficiently broad as to include ‘transmundanity’ in Schellenberg’s sense. 50

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

124

Agnosticism

could only achieve a natural good by being rightly related to a natural object/state of affairs. Consider the following options: [M non-T, A non-T, ST] natural and whose value is (merely) equal to or less than the highest natural good [M non-T, AT, ST] natural and whose value transcends any natural good [MT, A non-T, ST] transcendent and whose value is (merely) equal to or less than the highest natural good None of these is obviously and uncontroversially internally consistent. In this case, it may be that triple transcendence is the minimal requirement for making the idea of ST coherent. But suppose this idea of mere ST is internally consistent. The ultimist could also observe that a consequence of this is that there must be a non-doxastic faith stance which would have a still higher initial probability than ietsism.51 If this is so, then by the same logic which is supposed to justify the preference for ietsism over ultimism, former local agnostics should not (only) be ietsists, but adopt a still vaguer and more capacious position (let us call it R) which involves an attenuation even of ST. But it seems like this may well be so, since ietsism so defined will in its turn exclude a whole tract of actual and possible religious positions which fall under the umbrella term ‘religious naturalism’. This would mean that while ietsism has a higher initial probability than ultimism, in turn R would have a higher initial probability than ietsism. Consider a case like Leonard Angel’s ‘mystical naturalism’.52 Angel’s position can be formulated in Schellenbergian terms as follows: humans can achieve a transmundane (but non-transcendent) good by being related to a reality which is metaphysically and axiologically transmundane (but non-transcendent) By ‘transmundane’ Schellenberg means thoughts not restricted to the ‘mundane realm’, by which he understands ‘those aspects of human life and its environment to which just any mature human always has

51 This of course does not improve ultimism’s prospects, but it does lower ietsism’s. 52 See Angel (2002), pp. 317–38.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

125

quick and natural cognitive and experiential access.’ So an experience would count as transmundane if it was an experience, no matter how natural (e.g. an experience of unity with the totality of physical nature), to which mature human beings do not have quick and natural cognitive and experiential access. He also states that transmundane implies something ‘beyond’ or ‘more’, mentioning for instance Hindu claims about Brahman, or Taoist claims about the Tao as examples, and noting that the ‘precise force of the relevant use of “beyond” is elusive, and what such a reality is more than is difficult to specify.’53 In any case, transmundane is not co-extensive with ‘supernatural’ but is a broader category. Angel claims that mysticism and physicalism—which he defines as ‘the view that every mental or functional state is physically embodied’—are unproblematically compatible.54 Consider his construal of unitive mystical experience: The identification which is expressed in the statement ‘I am the first person’ is not fixed as the only form of identification, and that one can come to discover other forms of identification, such as ‘I am the many things; the first person is a proper part of me’, and ‘There is no me; there are only the many things’.55

He notes that the normal ‘ego perspective’ of mundane human experience presupposes ontological pluralism, that is, the metaphysical view that there is more than one object in the universe, which he defines more precisely as ‘The universe has many objects, and I am one of them, namely, this very person’, a claim he states is perfectly compatible with physicalism. Also, these other forms of identification do not involve diverging from an ontologically pluralistic metaphysical stance into ontological monism (‘there is exactly one object’) or ontological nihilism (there is no object at all), but they only involve ‘deny[ing] the identification of self with the first person’, and in two ways: (1) ‘see self-referring terms as having the composite whole, the universe, as their referent’ (= ‘I am the universe’); and (2) ‘deny that there is any extension to the self concept at all (= ‘There is no self, there is only the many objects’).56 In this way he maintains the metaphysical stance of ontological pluralism, but the ‘self placement’ is altered. This position he calls the 53 55

54 Schellenberg (2005), pp. 11–13. Angel (1994), p. 298. 56 Angel (1994), p. 298. Angel (1994), p. 299.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

126

Agnosticism

position of the ‘self-altered mystic’, which can take either of two forms: universal self mysticism (as for instance in neo-Vedanta) versus no-self mysticism (as for instance in Buddhism).57 In other words, this (among other things) gives Angel a means of offering an entirely naturalistic interpretation of unitive (monistic) mystical experiences, without offering a ‘reductive’ explanation of them (if ‘reductive’ is somehow meant to imply that they are in some respect illusory). Angel argues that it is not at all obvious how any of this is incompatible with physicalism, since even the eliminativist’s rejection of self-altered mystics’ propositions on the basis that they are couched in folk psychological language is not the kind of objection which hits the target, so that at least the no-self mystic (and, he suggests, perhaps even the universal mystic) could embrace the eliminativist’s account of what is happening.58 Certainly these experiences of identification with everything, or no-self experiences, and the propositions they give rise to count as ‘unusual perspectives and unusual ways of putting things’, and in that sense are transmundane, so these diverse phenomenologies of selfidentification do not require any reference to MT, but only transmundanity.59 Similarly, this kind of naturalistic mysticism does not imply axiological or soteriological transcendence, but only that ‘the best thing an individual can strive for qua monadic individual is steady, blissful awareness of Ultimacy, moment by moment.’60 Egorelease gives rise to the ‘joy of moral attitudes’ (naturally culminating in pure altruism), since it is ‘a consequence of seeing that every human being is a (proper) part of me.’61 Since for the purposes of a temporalist argument we only need to establish its epistemic possibility we do not need to worry about the details of Angel’s position beyond being able to indicate its prima facie plausibility. Can ietsism accommodate religious (?) positions like this, or would it only be captured by the more capacious religious 57

58 Angel (1994), p. 299. Angel (1994), p. 300. Angel (1994), p. 301. 60 Of course ‘ultimacy’ is not used in Schellenberg’s sense here. 61 Angel (1994), p. 305. This plainly excludes transcendent axiological and soteriological goods such as an afterlife which might atone for injustices etc. Cf. Schellenberg: ‘there is clearly no possible world in which a child whose life is filled with suffering and then cut short has already achieved all that “soteriological ultimacy” could sensibly be claimed to include for her’ (Schellenberg (2013c), p. 107). 59

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

127

framework proposition R? If only by the latter, and if Angel’s position is an authentically religious one, the erstwhile local agnostic should adopt non-doxastic faith in R. However, perhaps Elliott uses ‘transcendence’ in a sense which is sufficiently broad as to include transmundanity. This may be implicit in his appeal to transcendence ‘in the most epistemically humble and parsimonious way possible.’62 In which case, on this interpretation these objections will be moot, since one can imagine ietsism formulated so parsimoniously that it includes transmundanity at one extreme of the spectrum of religious positions it encompasses.63 If then by ‘soteriological transcendence’ we understand ‘soteriologically transmundane or transcendent’, this latter construal of an ietsist position does carve out an area within the religious possibility space which falls outside ultimism and is perhaps both internally consistent and the most general claim one could adopt without leaving religion. But conceivably the ultimist could insist that the requirement for a religious framework proposition to ‘find some place for powerful religious experiences’ is still much better satisfied by ultimism than by ietsism. In his Prolegomena, Schellenberg defines the tendency to ultimize as ‘a tendency to experience ultimizing (i.e. limit-removing) thoughts, or to engage in limit-removing behaviour, or else to feel limit-removing emotions’, and such thoughts relate to the objects of their concern as though they were ultimate, whether explicitly (i.e. in thought) or not.64 Certainly the ietsist and the ultimist can agree that a subset—on the total evidence, at least a subset, and on current evidence, a large subset—of mystical experiences exhibit limit-removing features of this kind. But the issue is whether this subset is the most representative of what mystical experience is essentially experience of, should it turn out that such experience is veridical. In defence of his tendencyto-ultimize criterion, which also allows him to say that a person is ultimizing even where she does not consciously and explicitly do so, 62

Elliott (2017), p. 109. I am grateful to Elliott for helpful clarification on this point in personal communication. 64 Schellenberg (2005), pp. 16–17. See also: ‘The most powerful religious experiences . . . positively encourage totalization or ultimization in religious thought. The awe, wonder, and other numinous states elicited are unwilling to rest content with any characterization falling short of unlimited richness’ (Schellenberg (2013c), p. 97). 63

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

128

Agnosticism

Schellenberg notes that it can even explain, for example, cases where religious persons deny that the God they worship has omniattributes. This is because their behaviour (worship), or religious emotions/experience, are ultimizing. Schellenberg also argues that if somebody failed to ultimize in any of these dimensions (cognitively, behaviourally, emotionally), then they should not be classed as religious, even if their behaviour externally resembles that of religious behaviour (e.g. somebody impersonating a priest), they are part of an institutional religion, or they self-identify—or are identified by others—as religious. He does not however rule out non-ultimistic forms of religion completely, noting that these, ‘if such there be’, could be thought of as a ‘second tier, reserved for later exploration.’65 This does acknowledge that some non-ultimistic religious possibility could turn out to be the case, and Elliott’s appeal to initial probability has force here too. But the question is whether a non-ultimistic religious framework proposition or propositions should be investigated concurrently with ultimism, or whether it should replace ultimism as the most reasonable nondoxastic faith stance. In both respects Schellenberg can still perhaps make a case that ultimism is the more rational option. But to return to religious experience, Schellenberg’s claims that much religious experience is fully ultimistic, and there is an ultimizing tendency in (all?) religious experience are open to at least the following objections: a) some mystical experiences do not have any ultimistic, but only transcendent, features; b) some mystical experiences only have transmundane features— for instance, pantheistic nature mystical ones, which merely have the physical universe as their object; c) humans may be unable to discern whether their experiences are of an ultimate or merely transcendent (or even transmundane?) object of concern.66 65 Schellenberg (2005), p. 34. To which he adds: ‘(perhaps on the conditions that further investigation into the concept of religion confirms their status as religious and that investigation into the first tier reveals nothing of enduring religious and philosophical value).’ 66 As Elliott has argued ((2017), pp. 108–9). He also argues (sketchily, as he admits) that although it is ‘intuitive’ that mystical experience is good evidence that its object is ‘utterly splendid, surpassing any sort of natural value’, this does not mean

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

129

In reply, an ultimist could simply bite the bullet and insist on the distinctiveness of an important set of religious experiences and how their distinctive phenomenology demands explanation. He notes that one reason for ‘taking something like ultimism more seriously even within purely religious contexts’ comes directly from ‘much’ religious experience, especially the most powerful kind, which carries a sense of an absolutely limitless richness that most makes me want to apply the term ‘religious’ to an experience. There has to be some term for experiences and ideas that want to burst all limits; what would it be if not ‘religious’?67

In other words, powerful ultimizing religious experiences are experienced as distinct from (a) transcendent or (b) merely transmundane ones and can be picked out from the others by a distinct term. The ultimist can point out that humans can and have reliably picked out this class of experiences as distinct from merely transcendent, or transmundane, ones, and—importantly—this is the case regardless of whether they consider the experiences veridical or not. In this respect, ultimism’s better fit with a large class of religious experiences would seem at first appearance at least to undermine the claim that ietsism can do a better job than ultimism in accounting for religious experience. This is because if any elaborated ultimism turned out to be the case, this would be unsurprising given the phenomenology of the ultimistic class of religious experiences. However, if any elaborated non-ultimistic religious possibility turned out to be the case, that is, within the larger ietsistic set,68 this would be surprising given the same phenomenology and fail to explain it. With respect to objection (3), Elliott has objected to Schellenberg’s appeal to religious experience on the grounds that unless it is already known ‘how an ultimate (or triply transcendent) reality will present itself ’, it is hard to say whether a person has been presented with it in

that ‘this value is ipso facto manifested as metaphysically significant (MT), intrinsically significant (AT), or extrinsically significant (ST)’ (p. 115, ft.34). 67 Schellenberg (2013d), pp. 277–8 (reply to Diller). 68 Ietsism also contains the subset of elaborated ultimistic religious possibilities. These of course would not be surprising given the ultimistic class of religious experiences. But adding a subset of elaborated non-ultimistic religious possibilities which would be surprising given the same phenomenology, would not put ietsism in a better position than ultimism to account for the data of religious experience.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

130

Agnosticism

such an experience. Since such a person would not know how to distinguish an experience of ultimate versus merely transcendent reality on account of ‘sheer imperceptiveness when it comes to transcendence’, humans are only justified in supposing that religious experiences point to transcendence simpliciter.69 However, to this objection an ultimist might simply appeal to trust in experiencers’ ability to discern differences (and researchers’ trust in the experiencers’ powers of discernment). If the objection is that religious experiences are equally evidential for the following two claims: Reality is metaphysically, axiologically and soteriologically at least transcendent70 Reality is soteriologically transcendent and at least metaphysically and axiologically transmundane then the ultimist can simply sit tight and reply that distinct types of religious experience can be distinguished from each other by their distinct phenomenologies. For instance, nature mystical experiences can be distinguished from other religious experiences on account of the felt sense of unity, love, etc., being of or with the natural world, rather than of something supernatural.71 These are distinct from ones which are (as Schellenberg says) ‘limit-transcending’, where the same qualities of love, unity, knowledge, etc., have a limitless feel about them. Also the ultimist could object that the ietsist would need to account for how, if her imperceptiveness claim is correct, experiencers seem to be clearly aware of the distinct phenomenologies of extrovertive and introvertive experiences but that nevertheless this discriminative ability—irrespective of the veridicality question— should somehow be discounted. Consider for instance Ann Taves’ distinction between types of specialness in her ‘building-block’ approach to religious experience.72 Taves’ approach to religious experience is a naturalizing one and supposes that ideal/absolute features are merely ascribed to experiences by humans rather than informing them about the nature of

69

Elliott (2017), p. 109. The options here range from MT, AT, ST to MU, AU, SU, and the logically possible intermediate positions. 71 72 See for instance Marshall (2005), pp. 24–7 and passim. Taves (2011). 70

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

131

ultimate reality.73 According to Taves’ model, religious experiences are a class of special experiences, but experiences can stand out as special by seeming ideal, both relatively or absolutely, or by seeming anomalous. Ultimizing experiences are (in Taves’ terms) the class of experiences which stand out as absolutely ideal, that is, as ‘no longer on a continuum with limited things, however special such things might be, but . . . fully set apart.’74 Given that there is a class of phenomenologically ultimistic religious experiences—whether or not they may turn out to be veridical—the ultimist could argue that although an elaborated ietsism is logically more likely to be true than an elaborated ultimism since it will be more inclusive, the members of the sub-class of nonultimistic religious possibilities within the ietistic set are less likely to be true than the ultimistic ones since as religious worldviews they lack an explanation for why an important class of religious experiences should be phenomenologically ultimistic. In other words, as briefly mentioned above, the scenario: when the total evidence is known, an elaborated non-ultimistic religious reality turns out to be the case would be surprising given the subset of ultimizing religious experiences which contraindicate such a religious reality, but unsurprising were an ultimistic reality to turn out to be the case. This is because on the face of it ultimism can explain the phenomenology of a widely distributed and culturally valorized class of mystical experiences which its experiencers naturally describe as ultimistic. It can also explain the following: why ultimistic descriptions of these experiences exist in human cultures, the cross-cultural prevalence of ineffability claims, and the class of counter-expectational experiences where the experience exceeded what the experiencer (perhaps with prior nonultimistic beliefs) anticipated. Unless the ietsist is prepared to abandon realism for a radical (neo-Kantian) contextualism and claim that such experience is constructed ‘all the way down’, it seems reasonable on the evidence to suppose that the ultimistic language is at least in part motivated by the phenomenology of the experiences themselves, which prompt experiencers to try to communicate about these using 73 Although Taves observes a methodological agnosticism about the worldview implications of this naturalizing approach. 74 Taves (2011), pp. 36–8.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

132

Agnosticism

ultimistic language. Of course this does not yet support a transcendent interpretation—it may be that the phenomenological distinctiveness of these experiences is due to natural factors—but it does require some explanation. Perhaps the ietsist could object that, for all we know, seemingly ultimizing religious experiences might be genuinely transcendent experiences but with a superaddition of illusion. Initially this seems promising, since there may be plenty of actual and future ways in which it could be imagined that mystical experiences are illusory. Therefore for all we know the sub-class of ultimistic religious experiences may be part veridical (qua revealing a transcendent reality) and part illusory (qua erroneously amplifying this veridical perception of transcendence into one of ultimacy). Suppose the following scenario: over an indefinite period of time into the future the scholarly and philosophical investigation of religious experience develops ever more sophisticated phenomenological reports,75 correlations with third personal (neurophysiological, etc.) evidence, detection of illusions, understanding of psychopathology, etc. Suppose also that at some future time x a scholarly and philosophical consensus emerges that the reports of merely transcendent experiences are largely or wholly free of illusion, but the reports of putatively ultimistic experiences are without exception found to contain one or more illusory features which account for the idealizing/absolutizing tendency as a (conscious or unconscious) human ascription. On this scenario, it would indeed by the case that the religious experiencers who had ultimizing experiences would not have known that they were experiencing a merely transcendent object. This seems like quite a promising objection to ultimism. Perhaps the ultimist can fall back on appeal to something like a principle of credulity,76 although of course ‘credulity’ will be the inappropriate word here, since we are dealing with non-doxastic faith rather than belief. Schellenberg speaks of religious experiences ‘all of us have had’ which he says can be called transcendent, ‘very important sorts of 75 In keeping with the earlier supposition that it is reasonable for this hypothetical agnostic to assume realism, I am presupposing that strong contextualism—a form of neo-Kantianism—is not in consideration here. Of course, if it is, the question of religious experience’s veridicality is moot. 76 I am grateful to John Schellenberg for suggesting in email correspondence the potential bearing of something like a principle of credulity in response to this sort of objection.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

133

human experience that might naturally lead one to consider the possibility of a religious dimension of life, and to think it worth pursuing’.77 He further suggests—and here I return to the third Schellenbergian reason for passing beyond pure scepticism alluded to briefly at the beginning of this chapter—that there is no reason why these hunches/experiences may not serve a ‘motivational function’.78 Of limit-removing religious experiences, he says that one may carry on in the train of thought that the beauty experience has begun by imagining that beauty is not shallow in the nature of things . . . [a]s Robert Nozick suggests . . . to neglect such promptings would only reveal a regrettable distrust in one’s positive responses to the world, and thus involve a ‘significant alienation from oneself ’.79

Granted that belief on the basis of such experiences is unjustified, one can nevertheless be faithful to one’s responses in the sense of imaginatively developing them where there is no reason not to. Although even for experiencers these experiences are not sufficient for belief, either as self-authenticating or as one kind of evidence within a larger cumulative argument for a religious reality, nevertheless, without a reason not to, arguably it is reasonable to remain faithful to one’s responses to the world. This sort of modified principle of credulity would operate at the level of faithfulness to one’s intuitions/hunches/ emotional responses to reality.80 In other words, powerful religious experience (of the sort supporting the ultimistic class of religious possibilities), while not providing us with grounds for religious belief, can motivate us to imagine that ultimism is true and act according to ultimistic principles. So we might say that for the person who has ultimistic experiences, it is reasonable to trust them in the sense of being open to them and letting them lead our imaginations and our propositional faith. Discussing the aim of ‘imaginative fulfilment’, Schellenberg notes that humans have an imagination which ‘is sufficiently spacious and grand to hold the idea of the Ultimate . . . and also is bent towards pursuing an expanded understanding capable of embracing this idea 77 Schellenberg (2009), p. 180. Of course the ietsist or religious naturalist will contend the claim that ‘all of us’ have had ultimistic experiences. 78 Schellenberg (2009), p. 51. 79 Schellenberg (2009), p. 150 (the reference is to Nozick (1989), p. 52). 80 Of course the ietsist and religious naturalist can develop equivalent arguments for transcendent or transmundane religious experiences.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

134

Agnosticism

ever more fully’; and if we add to this that ultimism is not ‘obviously false’ and ‘[i]t cannot be ruled out that we are destined for ultimate things, and that the imaginative activity in question will turn out to be fundamentally realistic, only hastening our arrival at that destination’, ultimism need not be abandoned, unless perhaps there was already very good cumulative evidence that the ultimistic features of human experiencing and cognition were clearly illusory, and some reason to suppose that this evidence was of the sort which was relatively futureproof.81 Suppose then that something like a principle of credulity can be revived in this way in a faith key, as it were. The ultimist could argue that given everything we humans do not yet know, these responses may belong to our unavoidable reactions to the world, and even though we have reasons to think there are or may in the future be defeaters to beliefs formed on the basis of them, there are no reasons to suppose they shouldn’t motivate a faith stance. Indeed, Schellenberg would argue, not having faith in them would represent an unjustified distrust in oneself and one’s own responses without any good reason for it.82 (This is also the reason why pure scepticism is not more rational than ultimistic non-doxastic faith: in addition to being epistemically on a par with, rather than superior to, ultimism— since the ultimist also looks out for arguments which may reveal ultimism’s falsehood—it would represent an irrational distrust in one’s own unavoidable reactions to the world.) But suppose that truth-oriented considerations from religious experience are not decisive in swinging the argument for the rational preferability of ultimism or ietsism one way or the other. There is also the further consideration that, for all we know, humans may need vast amounts of time—perhaps pushing up to the maximum span of time available to them—to reach reliable consensus on ultimate reality. Given this possibility, the choice to investigate non-ultimisms as well as ultimisms at the same time rather than send them to ‘the back of the line’ may slow down religious enquiry such that tragically consensus is reached much later or (conceivably) not at all.83

81

82 Schellenberg (2009), pp. 114–15. Schellenberg (2009), pp. 180–1. The ietsist could also of course turn this sort of (highly speculative) argument in her favour. But the point is simply that ietsism’s having prior probability on its side does not automatically translate into its being preferable as a research programme for religious enquiry. 83

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

135

However, supposing one can imagine reasonably plausible ways in which it could meet such an objection, ietsism may be beset by a further and more substantial difficulty, that it is difficult to bring it ‘in focus’ as it were as an object of non-doxastic faith, and consequently as a religious guide to behaviour.84 In this sense ultimism may yet be the most general claim one could practically adopt as a non-doxastic faith stance.85 Elliott notes that ietsism is just like ultimism in containing an extremely large set of religious disjuncts, but—as is also the case with ultimism—a person has non-doxastic faith in ‘the disjunction of possibilities itself, not any particular disjunct’, that is to say, in bare generic ietsism, unelaborated in any way.86 Here is the difficulty with bringing ietsism ‘in focus’ as an object of non-doxastic faith. The faith-based ietsist must put her faith in such a superset of disjuncts, namely, those disjuncts which make up the conceptual possibility space of generic religious framework propositions of which ultimism is a species. Unlike the diversity within the set of ultimistic religious possibilities into which generic ultimism is resolvable, the diversity within generic ietsism is much greater. It will include all internally consistent generic religious framework propositions imaginable, ranging from those exhibiting mere transmundanity in one or more of Schellenberg’s dimensions (metaphysical, axiological, soteriological) at one extreme, to those exhibiting ultimacy in all three of the same dimensions at the other, with propositions exhibiting transcendence lying between these extremes.87 One can quibble about the appropriate taxonomy for these framework propositions, but the important point is that ietsism is not coordinate with these generic religious framework propositions, but is super-ordinate to them, as their superset. Therefore a curious property of ietsistic non-doxastic faith is that it includes everything from the barest religious naturalism (if it does) at one end of the religious

84 Schellenberg gestures to this problem in his reply to Diller. Schellenberg (2013d), p. 276. 85 This is distinct from ietsism’s prospects as a research programme for religious enquiry. But even here ietsism is not clearly preferable or free of difficulties. There is for instance the question of the place of religious naturalism: if the latter is included within ietsism, then the programme should be exhaustive; if it isn’t, it will be vulnerable to the initial probability objection which it itself levels against ultimism. 86 Elliott (2017), p. 112. 87 Examples would include religious naturalism at one extreme; triple transcendence in the intermediate range; and ultimism at the other extreme.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

136

Agnosticism

spectrum, to ultimism at its other end. In order to visualize the difficulty of bringing the object of ietsistic faith—which traverses transmundane, transcendent, and ultimate religious possibilities— into focus, consider for instance the following analogy: an actor at an awards ceremony chooses to have faith that she will secure the first, second, and third prizes (i.e. three simultaneous non-doxastic faith commitments). This would appear to necessitate a subjective superimposition of three ‘zeals’ or imaginative challenges distinct in quality and degree, as well as (in the case of her attitude towards the second and third prizes) two admixtures of sadness or disappointment (relative to the first prize being out of reach).88 As well as the difficulty of getting the object of ietistic faith into focus, there is also the additional difficulty: a commitment to ultimism plausibly has different behavioural implications than faith in (say) generic religious naturalism, or triple transcendence; there will be differences in human response to limited versus unlimited transcendent realities, expressed in different intensities of worship and different behaviours. For instance, worship of a limited deity known or supposed to be imperfect—or of nature itself—will differ from worship of a deity known or supposed to be the Perfect Being. Similarly, faith in a merely transmundane or transcendent religious framework proposition will plausibly elicit different human religious responses than faith in an ultimistic religious framework proposition. To use Schellenberg’s examples again: faith in the latter may elicit more reverential behaviour and more risk-taking behaviours for the sake of the good, etc.; but faith in the former may elicit these behaviours to lesser degrees, if—in the case of transmundanity—at all. In the case of ultimism, the elaborated ultimisms into which it can be resolved will all be of the same type—however else they differ, they will all exhibit metaphysical, axiological, and soteriological ultimacy—and thus elicit the same type of worshipful attitude and behaviour generically speaking. By contrast, in the case of ietsism, the elaborated forms, although all involving transcendence (or at least transmundanity?), will fall into very distinct types of elaborated religious possibilities: ultimism, generic religious naturalism, triple transcendence, etc., each type plausibly eliciting different intensities of worship and different behavioural responses. The ietsist must 88 That is, on the assumption that it would be objectively saddening to discover that something less than ultimism was the case, when one had previously had faith in ultimistic as well as transcendent and transmundane religious possibilities.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

137

therefore explain how a person can have non-doxastic faith in a superset of this sort, since each set of religious possibilities within the superset will presumably elicit a distinct quality and degree of commitment, which would seem to involve a sort of simultaneous subjective superimpositioning of multiple distinct non-doxastic faith commitments. How can such a person, who must somehow simultaneously have multiple non-doxastic faith commitments to quite different generic religious framework positions, follow the (ideal) behavioural implications of only one of them—for instance (in the case of the ultimistic set), to aim at an ‘infinite’ quality in human behaviour, say, unconditional love or limit-transcending awe, rather than (in the case of some non-ultimistic set) merely greater than normal love or awe, etc.?

V. CONCLUSION The formerly local agnostic, I have suggested, faced with the new conceptual territory for ‘agnosticism’ opened up especially by Schellenberg’s work, and other contributions within what Diller has helpfully termed the religious pluralism tradition, will not need to become a pure religious sceptic. Nor need she suppose that ietsism is the preferable research programme for religious enquiry, although admittedly ietsism must be a serious candidate for such a research programme. But with respect to non-doxastic faith I have suggested that notwithstanding important objections, it may remain more appropriate for her to be a ‘new agnostic’ in the Schellenbergian sense cited in the introduction to this chapter, namely, by embracing ultimistic non-doxastic faith. This is because ietsism may be difficult to bring ‘into focus’ as a non-doxastic faith stance, and consequently be unviable as a religious guide to behaviour. In this respect, of the two, ultimism may remain the most general religious claim which can be adopted as a non-doxastic faith stance, since (even setting aside the additional pragmatic considerations in its favour) it is perhaps the most general one which can be brought into focus and serve as a viable religious guide to behaviour.89

89 I would like to express my gratitude to James Elliott and John Schellenberg whose comments in email correspondence have greatly helped the formulation of my arguments.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

138

Agnosticism BI B LI OGR APHY

Angel, Leonard. Enlightenment East and West (Albany: SUNY, 1994). Angel, Leonard. ‘Mystical Naturalism’. Religious Studies 38, no. 3 (2002), pp. 317–38. Byrne, Peter. Prolegomena to Religious Pluralism: Reference and Realism in Religion (London: Macmillan, 1995). Diller, Jeanine. ‘The Conceptual Focus of Ultimism: An Object of Religious Concern for the Nones and Somes’. Religious Studies 49, no. 2 (2013), pp. 221–33. Draper, Paul. ‘Seeking but Not Believing: Confessions of a Practicing Agnostic’. In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul K. Moser (Cambridge: Cambridge University Press, 2002), pp. 194–294. Draper, Paul. ‘Faith without God: An Introduction to Schellenberg’s Trilogy’. Philo 14, no. 1 (2011), pp. 59–65. Draper, Paul. ‘Atheism and Agnosticism’. The Stanford Encyclopedia of Philosophy (Fall 2017 edn), edited by Edward N. Zalta. https://plato. stanford.edu/archives/fall2017/entries/atheism-agnosticism/ Elliott, James. ‘The Power of Humility in Sceptical Religion: Why Ietsism is Preferable to J. L. Schellenberg’s Ultimism’. Religious Studies 53, no. 1 (2017), pp. 97–116. Forrest, Peter. ‘An Examination of John Schellenberg’s Austere Ultimism’. Sophia 52, no. 3 (2013), pp. 535–51. Hick, John. An Interpretation of Religion: Human Responses to the Transcendent (Yale: Yale University Press, 2005). Insole, Christopher J. The Realist Hope: A Critique of Anti-Realist Approaches in Contemporary Philosophical Theology (Aldershot: Ashgate, 2006). Lehe, Robert. ‘A Critique of Peter Byrne’s Religious Pluralism’. Religious Studies 50, no. 4 (2014), pp. 505–20. Le Poidevin, Robin. Agnosticism: A Very Short Introduction (Oxford: Oxford University Press, 2010). Marshall, Paul. Mystical Encounters with the Natural World: Experiences and Explanations (Oxford: Oxford University Press, 2005). Nozick, Robert. The Examined Life: Philosophical Meditations (New York: Simon & Shuster, 1989). Rottschaefer, William A. ‘Schellenberg’s Evolutionary Religion: How Evolutionary and How Religious?’ Religious Studies 52, no. 4 (2016), pp. 475–96. Schellenberg, J.L. Prolegomena to a Philosophy of Religion (Ithaca, NY: Cornell University Press, 2005). Schellenberg, J.L. The Wisdom to Doubt: A Justification of Religious Skepticism (Ithaca, NY: Cornell University Press, 2007).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

‘New Agnosticism’

139

Schellenberg, J.L. The Will to Imagine: A Justification of Skeptical Religion (Ithaca, NY: Cornell University Press, 2009). Schellenberg, J.L. ‘My Stance in the Philosophy of Religion’. Religious Studies 49, no. 2 (2013a), pp. 143–50. Schellenberg, J.L. ‘In Praise of Austerity: A Reply to Forrest’. Sophia 52, no. 4 (2013b), pp. 695–700. Schellenberg, J.L. Evolutionary Religion (Oxford: Oxford University Press, 2013c). Schellenberg, J.L. ‘Replies to My Colleagues’. Religious Studies 49, no. 2 (2013d), pp. 257–85. Schellenberg, J.L. ‘God for All Time: From Theism to Ultimism’. In Alternative Conceptions of God, edited by Andrei Buckareff and Yujin Nagasawa (Oxford: Oxford University Press, 2016), pp. 164–79. Taves, Ann. Religious Experience Reconsidered: A Building-Block Approach to the Study of Religion and Other Special Things (Princeton: Princeton University Press, 2011).

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

5 Hegel, Wittgenstein, and the Question of Agnosticism Gavin Hyman

I Hegel and Wittgenstein may be thought unlikely philosophical interlocutors in a volume on agnosticism. This is because agnosticism—as it was originally conceived and as it has subsequently been understood—is predicated on the very epistemology and anthropology that both Hegel and Wittgenstein reject. Markedly heterogeneous though their projects are, they at least share a rejection of the dominant philosophical paradigm of modernity, wherein the ‘knowing’ subject—from a position of detachment—observes and describes the world in order accurately to reflect the ‘reality’ that it sees passing before it. This knowing subject is ‘abstracted’ from the world, ostensibly in the interests of neutrality or objectivity, but with the result that a ‘gap’ is installed—which in some cases becomes a chasm— between the subject and the world, between the knower and the known. Since Descartes—who effectively inaugurated this model— the philosophical task has been conceived as to how to ‘bridge’ this gap so as to avoid it becoming philosophically disabling. The history of modern philosophy in the West may be conceived as a series of various attempts precisely to bridge this gap. But there has also been an equally persistent strain within philosophy that concludes that this gap is simply unbridgeable. From Descartes’s deceiving demon, to Hume’s outright scepticism, to Kant’s conception of Gavin Hyman, Hegel, Wittgenstein, and the Question of Agnosticism In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0006

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

141

the noumenon, there has been a deep-seated suspicion of the very possibility of the mind being able to represent reality as it is in itself. Agnosticism belongs in this philosophical tradition of unknowing, this tendency to posit an irreducible gap between the knowing subject and the reality it observes. But when Thomas Huxley and his associates coined the new term ‘agnosticism’ in 1870, they did so in order to distinguish their disposition from other contemporary forms of unknowing. First, they wanted to distinguish agnosticism from more generalized forms of scepticism. Huxley and his associates were not necessarily sceptical about the possibility of knowledge of the physical and natural world, but they wanted to confess an a-gnosis regarding the possibility of transcendent truths, in particular, the question of the existence of God. Second, they wanted to distinguish themselves from atheists, partly for pragmatic and political reasons, but, more importantly, because atheists claimed to know too much, in that they were confident that the gap between the knowing subject and reality could be bridged and, having bridged it, were confident that God did not exist. Huxley and the new ‘agnostics’ believed there to be no warrant for such confidence, and, in this respect, they moved closer to the scepticism that in other respects they disavowed. We can thus see the extent to which agnosticism is contingent upon a specific epistemology which, in turn, has a particular understanding of the knowing subject at its heart. Indeed, this is so evidently the case that we may well wonder to what extent agnosticism would even be conceivable in a quite different philosophical context. In other words, if agnosticism is one answer to a specific philosophical question (regarding how the supposed ‘gap’ between knower and reality might be ‘bridged’), then what would become of it if this question were itself shown to be a misguided or false one? All of which is to bring us to Hegel and Wittgenstein. For all their undoubted differences, they would nevertheless agree that this is indeed a misguided question. For them, the human being is not a detached observer or knower of reality, but is rather an intrinsic part of, and submerged within, that reality. There is no external vantage point for finite human beings from which reality might be seen, surveyed, and represented. We are always immersed within the very reality from which we misguidedly attempt to abstract ourselves. Our language, outlooks, and forms of life do not ‘represent’ a reality ‘out there’, but are ways of responding to, and negotiating our ways around and within, our reality—a reality within which we are fully embedded.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

142

Agnosticism

Hegel and Wittgenstein would not only agree in their rejection of this modern epistemological model, but they would further agree that it produces distorted ways of asking and answering philosophical questions. In the case of God, for instance, it inevitably takes the form of the knowing subject asking whether it is reasonable to suppose that a being, person, spirit, or entity called God ‘exists’ in ‘reality’. For both Hegel and Wittgenstein, this is a deeply problematic question to ask, riven as it is with all kinds of problematic philosophical assumptions. The result is that whether one replies affirmatively, negatively, or agnostically, one cannot avoid the implication that one is being led astray. Whichever reply is proffered entails one being mired in confusion and distortion. It follows from this that Hegel and Wittgenstein’s thought would not only preclude agnosticism, but also theism and atheism—insofar as they are conceived to be answers to the problematic statement of the question presented above. But what if we were to follow Hegel and Wittgenstein in abandoning this modern epistemological model altogether? Could questions relating to theism, atheism, and agnosticism still be asked? If they could, in what ways would these questions and conceptions be transformed? In recent years, various thinkers have developed interesting answers to these questions, which are suggestive of qualitatively different understandings of theism and atheism from those conventionally developed by modern epistemology. Rowan Williams and Andrew Shanks, for instance, have asked what it would mean to take Hegel seriously as a theological thinker, and what kind of theism begins to emerge as a result.1 Slavoj Žižek, in contrast, has insisted on an atheistic reading of Hegel, albeit one in which Christianity remains central, and has emerged with a form of atheism quite unlike anything dreamed of by the so-called ‘New Atheists’.2 Fergus Kerr’s pioneering and now classic study asked how one might pursue theology after Wittgenstein, and how, in consequence, theism might be conceived otherwise than has conventionally been done by modern metaphysics.3 There have even been some tentative gestations considering what implications Wittgenstein’s philosophy might have 1 For Rowan Williams’s explicit discussion of Hegel and theology, see Williams (2007), chapters 2 and 3, although the Hegelian influence on his thought is evident throughout his writings generally. For Andrew Shanks, see, for instance, Shanks (2011) and (2014). 2 See Žižek (2012) as well as his work more generally. 3 See Kerr (1997 [1986]).

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

143

for a new conception of atheism.4 But if new forms of theism and atheism have been explored after Hegel and Wittgenstein, what of new forms of agnosticism? How might agnosticism be re-conceived in ways that are cognizant of Hegel and Wittgenstein’s strictures against modern epistemology? To what extent are there agnostic strains intrinsic to their own thought? Clearly, such a re-conceived agnosticism will not relate to a hypothesis concerning the existence of a being called God. But it could perhaps be re-conceptualized in accordance with the meaning of the Greek words from which the term is derived—as a generalized unknowing; and it may also be fruitful to consider how this could relate further to the question of God. These are, I believe, inherently interesting questions, for Hegel and Wittgenstein are not thinkers who are usually associated with a disposition of ‘unknowing’. Hegel is often thought to aspire to ‘Absolute’ knowledge, a philosopher with pretensions to ‘know everything’, whose System is supposed to leave nothing out, and to leave nothing ‘unknown’. Wittgenstein, on the other hand, is often thought to repudiate metaphysics in any form, insisting that ‘nothing is hidden’ and ‘all is open to view’; there is nothing, in other words, to be agnostic about. But these are, of course, misleading caricatures, and in what follows I want to explore the ways in which a certain kind of agnosticism, of unknowing, permeates the thought of them both, albeit in different ways. As one would expect of such heterogeneous thinkers, the way in which a certain kind of agnosticism manifests itself in each thinker is strikingly different, both qualitatively and quantitatively, but these differences will themselves be instructive for any consideration of an agnostic disposition in a post-metaphysical context.

II For Hegel (as indeed for Wittgenstein), modern Western philosophy, at least since Descartes and as exemplified by Kant, was predicated on a mistake. This mistake was the false separating out and dividing up of that which is in reality conjoined. Pre-eminently, this was enacted 4 Although not greatly developed, the seeds are sown in Phillips (1996) and Hyman (2010), pp. 182–5.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

144

Agnosticism

in the separating out of the self, the subject, the ‘I’ that ‘knows’, from that which is ‘known’—reality, the world, that which is out there to be known. But this false separation gives rise to a whole host of other derivative separations—of dualisms—by which it might be said that modern philosophy is defined. These dualisms reach their apotheosis in Kant, where a non-negotiable separation between the ‘knower’ and the noumenon is installed. While reality may be known by us as the phenomenal world, what reality is ‘really’ like, in itself, apart from us, is forever inaccessible to us. In this sense, there is an absolute split between the subject and its world. For Hegel, this split between the knowing subject and reality, and the dualistic philosophical framework to which it gives rise, is problematic in at least two ways. First, it is a false picture of how human beings are related to the world in which they find themselves. When human beings attempt to ‘reflect’ or ‘represent’ their world, they are immediately faced with a limit: the fact that they are themselves a part of, and embedded within, that which they are attempting to represent. The act of representation can only be enacted by assuming the fiction of the subject’s detachment from the world, a fiction that immediately disfigures and distorts, and in this sense it is wrong, an error or a mistake. But it is not, of course, an error in the sense of being an incorrect representation of reality, which may be succeeded by a correct representation, for it is the very act of subjective representation that is here being questioned. It is not that the picture fails to ‘correspond’ to how reality is in itself, but rather that the picture is inconsistent or out of joint with how we experience life. In this sense, another division—between thought and experience, philosophy and life—is likewise being questioned, and the prospect is therefore being raised of a quite different understanding of what it means to ‘think’ and therefore what it means to ‘know’ (and not to know). Furthermore, as is well-known, Hegel’s understanding of an ‘error’ or ‘mistake’ is likewise re-conceived. On the dualistic understanding, an error was simply to be discarded as worthless, ‘as when we say of something that it is nothing, or is false, and then, having done with it, turn away and pass on to something else’.5 Rather, for Hegel, an error participates in an exchange, a conversation, or an educative process; it is part of a wider movement of advance.

5

Hegel (1977), p. 19.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

145

The second way in which the split between the knowing subject and reality is problematic for Hegel—beyond being philosophically false—is that it is experientially damaging for living human beings. The gap between the knower and reality—especially when it is hypostatized in the manner effected by Kant—cuts individuals off from their environment, prevents them from being ‘at home’ in the world, with the result that they become alienated from their own surrounding reality. In a sense, therefore, a falsely dualistic understanding of reality actually produces suffering or an ‘unhappy consciousness’; a false philosophy actually has existential consequences. Once again, therefore, we see a Hegelian deconstruction of the distinctions between thought and experience, philosophy and life. For Hegel, human beings are not just ‘knowing’ subjects, but are also ‘subjects of desire’, as Judith Butler has put it.6 This means, of course, that the prescribed therapy is not simply a matter of ‘correcting’ one’s philosophy, of acquiring a new worldview, but of enacting a new manner of being in the world. In this sense, the therapy is as much experiential as it is intellectual. Hegel gives some indications of what this entails when he says of the ‘old’ way of thinking that ‘instead of getting involved in the real issue, this kind of activity is always away beyond it; instead of tarrying with it, and losing itself in it, this kind of knowing is forever grasping at something new; it remains essentially preoccupied with itself instead of being preoccupied with the real issue and surrendering to it.’7 So at the outset of the Phenomenology we already get a sense of the key difference between the metaphysical way of doing philosophy and the new one being commended: in the former, the self is active— grasping and preoccupied with itself, while in the latter, the self cultivates a certain passivity—it loses itself and surrenders itself: ‘the individual must all the more forget himself . . . Of course, he must make of himself and achieve what he can; but less must be demanded of him, just as he in turn can expect less of himself, and may demand less for himself.’8 What we see here, therefore, is that Hegel ‘dissolves and maintains the ego in the world’,9 and what is being commended is a complete re-orientation of the conception of the ‘self ’ and its relation to what is ‘other’. In this sense, Hegel appears to anticipate Levinas’s later re-orientation of the subject which he characterized as 6 8

7 Butler (1987). Hegel (1977), p. 3. 9 Hegel (1977), p. 45. My emphasis. Rose (2009 [1981]), p. 212.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

146

Agnosticism

a move away from a nominative self—which asserts and projects—to an accusative self—which is itself posited and addressed (even if Hegel would have had parted company with Levinas in the latter’s characterization of the ‘other’). When the knowing subject asserts itself by grasping at what it takes to be truth, it forgets that what it asserts is actually a reflection of itself, and vice versa. So when the knowing self asserts its own fixity and stability, this means that what it posits has the same fixed and stable character, and it is this fixity, whether in the knower or what is known, that is problematic, and that occludes an openness to truth; in order that truth may appear, fixity must give way to fluidity: Thoughts become fluid when pure thinking, this inner immediacy, recognizes itself as a moment, or when the pure certainty of self abstracts from itself—not by leaving itself out, or setting itself aside, but by giving up the fixity of its self-positing, by giving up not only the fixity of the pure concrete, which the ‘I’ itself is, in contrast with its differentiated content, but also the fixity of the differentiated moments which, posited in the element of pure thinking, share the unconditioned nature of the ‘I’.10

What we see from this, therefore, is that the surrendering of the fixed ‘I’, the knowing subject, entails not only a forgetting of itself, but also a losing and surrendering of that which the knowing subject posits. There is, in effect, therefore, a double movement of ‘unknowing’, what we might characterize as a double movement of ‘agnosticism’. There is a surrendering of the subject’s self-certainty, of its own knowledge of itself, and this, in turn, entails a surrendering of that which the subject claims to ‘know’ through its own assertive act of positing. Through this double agnostic movement, the fixity of the distinction between the knower and the known subsides, as well as the fixity of the knower and what is known in themselves. These movements open up the possibility for what Hegel calls the fluidity of thought, which is itself a precondition of enlightenment. We shall return to this question of enlightenment in a moment, but we should pause first to consider further the nature of this agnostic movement in Hegel, and in particular its relation to more conventional forms of agnosticism. There are at least three ways in which Hegel’s agnosticism is distinguished from more conventional forms. 10

Hegel (1977), p. 20.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

147

First, agnosticism as traditionally understood does not extend its unknowing to the knower itself. The subject, in other words, remains unquestioned. That which is posited is subject to doubt or unknowing, but that which is doing the positing is not similarly doubted. In this sense, Hegel’s agnosticism may be seen as a more radical form of unknowing. Not content simply to doubt knowledge claims, Hegel insists on extending that doubt to the subject itself who makes these claims. Second, Hegel’s agnosticism or ‘unknowing’ is not applied to a particular form of positing (‘the existence of God’) which fails to fulfil certain conditions which other forms of positing (knowledge of the natural world) do fulfil. Rather, the ‘unknowing’ or ‘path of despair’ questions the activity of positing itself. As Gillian Rose has characterized it, it is a despair which questions representation as such, and which seeks ‘conscious insight into the untruth of phenomenal knowledge’, into the ‘so-called’ naturalness of the representation. Representation which appears as knowledge, as truth, cannot be true, but must, by definition be misrepresentation . . . . This path is self-perficient, self-completing, because it is more radical than mere doubt, and because it presents the ‘complete’ forms of ‘untrue consciousness in its untruth’.11

Hegel’s agnosticism is again therefore more radical than conventional forms because it doubts not just this or that claim to knowledge, nor indeed every claim to knowledge, but the very mechanism of representation itself. Third, in a conventional form of agnosticism or indeed in a generalized form of scepticism, which is distinguished from Hegel’s unknowing in the ways just described, the knowing subject finds itself ‘stuck’ without any possibility of advance or way out of the deadlock. This conception of agnosticism or unknowing as a telos is, of course, an effect of the ‘fixity’ of the knowing subject and the ‘fixity’ of that which it posits, a fixity which Hegel seeks to dissolve. Correspondingly, therefore, Hegel’s agnosticism is not an end-point or an unknowing achieved for its own sake, but a prelude to a greater knowing. But the key point here is that this greater knowing, this form of enlightenment, does not simply cancel out and annul the preceding agnosticism. To understand it thus would be consistent with the clichéd reading of Hegel, which so many recent commentators have 11

Rose (2009 [1981]), p. 162.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

148

Agnosticism

been at pains to question. This clichéd reading sees all negation in Hegel as being penultimate, as always being annulled and gathered up into a higher unity, so that negativity is always ultimately subordinated to positivity. But it is far from clear that this is the case. We have already noted the importance for Hegel of the knowing subject surrendering itself, emptying itself out, and so forth, and it follows from this that the enlightenment which both constitutes and succeeds the insight into the ‘untruth of phenomenal knowledge’ is something that ‘befalls’ the human being, something to which the person is susceptible, rather than something the knowing subject achieves by its own efforts and rational processes. In this sense, the enlightenment is more akin to a religious revelation than to philosophical ratiocination. For Hegel, this period of transition was one that was coming to light, making itself known, in his own time. As he put it, The frivolity and boredom which unsettle the established order, the vague foreboding of something unknown, these are the heralds of approaching change. The gradual crumbling that left unaltered the face of the whole is cut short by a sunburst which, in one flash, illuminates the features of the new world.12

It is this flash, this illumination, which constitutes enlightenment, and which means that we do not remain stuck in deadlock in the previously articulated unknowing. This ‘sunburst’ which ‘illuminates’ in ‘one flash’ is what Žižek has characterized as a manifestation of the ‘parallax gap’ where in one sense all remains the same, but with a shift of perspective and a new angle of seeing, all is transformed and seen anew. The truth of phenomenal knowledge, which was previously only seen in its untruth, is suddenly revealed. But there is a sense in which this revelation conceals as much as it reveals. In the passage that immediately succeeds the one just quoted, Hegel continues: But this new world is no more a complete actuality than is a newborn child; it is essential to bear this in mind. It comes on the scene for the first time in its immediacy or its Notion. Just as little as a building is finished when its foundation has been laid, so little is the achieved Notion of the whole the whole itself.13

12

Hegel (1977), pp. 6–7.

13

Hegel (1977), p. 7.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

149

So when enlightenment has come forth as a ‘sunburst’, previous configurations of consciousness which had hitherto understood themselves as being ‘true’ or ‘false’, or as at one time ‘true’ and later ‘false’, are now seen for what they really are—stations along a path which ultimately makes up this larger whole, and it is this larger whole which itself constitutes Truth. And yet, this insight takes us little further than a foundation does in relation to a building. There is still much that is ‘unknown’ for it is yet to be realized: ‘this path is the conscious insight into the untruth of phenomenal knowledge, for which the supreme reality is what is in truth only the unrealized Notion’.14 So the revelation, the ‘sunburst’, does not do away with ‘unknowing’, but actually reinstates it, this time at a higher level, at the level of the Notion, the level of absolute knowledge. That Hegel should be agnostic or insist on the unknowability of the Notion is a necessary corollary of his renunciation of a ‘God’s eye view’, that is to say, his repudiation of the abstract knowing subject which ‘observes’ the world and has pretensions to describe that world as a whole. Because we are always immersed in the world, placed in it at a particular point in time, within a specific moment of the unfolding of the Notion, the perspective of the Absolute is inaccessible to us, something which is ‘unknown’ to us. But it is important to emphasize the nature of this agnosticism; Hegel is not agnostic about whether the Notion or the Absolute ‘exists’. This would be a meaningless question because the Notion is unfolding before our eyes; indeed, more than that, we are ourselves a part and a manifestation of this very unfolding. But what cannot be seen is the totality of this unfolding as a whole, and in this sense the unknowing is irreducible for finite human beings within time. But in principle, nothing is hidden behind a veil, nothing is like Kant’s noumenon, forever and a priori inaccessible. In this sense, Hegel proclaims (like Wittgenstein) that ‘nothing is hidden’. This is why Hegel, describing his task in the Phenomenology, says that ‘an exposition of how knowledge makes its appearance will here be undertaken’.15 So the nature of Hegel’s task is not one of reasoning, asserting, and arguing, but one of looking and seeing. As he puts it, we do not need to import criteria, or to make use of our own bright ideas and thoughts during the course of the inquiry; it is precisely when 14

Hegel (1977), p. 50.

15

Hegel (1977), p. 49. My emphasis.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

150

Agnosticism

we leave these aside that we succeed in contemplating the matter in hand as it is in and for itself . . . all that is left for us to do is simply to look on.16

When we look on, we see the repeated patterns of success and failure, of knowing and unknowing, of divisions and dichotomies that are both necessary and destructive. And what these patterns suggest or point to is a unified totality of consciousness, its ‘essence’ or Notion, or absolute knowledge, in which all these concrete particularities are gathered up. The absolute is knowable in the sense that it is unfolding before, around, and within us. But it will not be realized, cannot be stated, other than as a whole, as the totality of consciousness, and insofar as this is so, it is unknowable. We see in Hegel, therefore, the mutual interplay of knowing and unknowing, of knowledge and agnosticism. The knowability of the world is only intelligible in terms of and in the context of the unrealized Notion of which it is itself the manifestation. Contrariwise, the unknown Notion is only intelligible in terms of its own manifestation and unfolding in the known world. The interplay between known and unknown is ceaselessly in motion until the final realization of the Absolute, wherein the very distinction between known and unknown will be overcome. This interplay between the simultaneous knowing and unknowing of the absolute has been well-expressed by Gillian Rose, when she says: The absolute is the comprehensive thinking which transcends the dichotomies between concept and intuition, theoretical and practical reason. It cannot be thought (realized) because these dichotomies and their determinations are not transcended. Once we realize this we can think the absolute by acknowledging the element of Sollen [projection of an ought] in such a thinking, by acknowledging the subjective element, the limits on our thinking the absolute. This is to think the absolute and to fail to think it . . . .17

This is a good characterization of what I am here taking Hegel’s agnosticism to be—the necessity of the failure to think the absolute. And yet this is an agnosticism born out of our very thinking of the absolute, out of our knowing the absolute—not as such but in its manifestation in the repeated statements and failures of our thought. Rose’s suggestion is that to attempt to state it as such—to attempt to 16

Hegel (1977), p. 54.

17

Rose (2009 [1981]), p. 218. My emphasis.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

151

overcome the agnosticism—would be to reduce it to the contradictions and divisions of our concrete thought, and thus to destroy it as such. Thus, the agnosticism must be preserved, even if we can see the absolute all around us as manifested in the movement between the concrete ‘moments’ of consciousness in its actuality. The absolute is ‘conceivable but impossible’,18 and this is what Hegel’s agnosticism is—not just a limit of thought, but also a precondition of thought.

III Wittgenstein is not often thought to be a philosophical accomplice of Hegel, but in fact he is grappling with similar problems and engaging with the many of the same issues, albeit emerging with quite different solutions. Wittgenstein shares all of Hegel’s concerns about the detached knowing subject which has been paradigmatic for modern philosophy. This has been well-expressed by Fergus Kerr: Regarding ourselves as detached observers of the passing scene encourages us to treat language as representation of reality, and thought as mirror-image of the world. However natural and venerable it may be to think along such lines, it is yielding to metaphysical antipathy to life. Wittgenstein challenges this entire tradition in a sentence: ‘Would it be correct to say our concepts reflect our life? They stand in the middle of it’ . . . We cannot catch reality in our net because it is the bustle of life of which we are a part. The idea that thought, or language, represents the reality ‘out there’, however natural and intuitively appealing, draws us easily into supposing that the whole hurly-burly may be treated as just one more object to be represented, as if we could get outside it to view it from somewhere else. This is one more way in which the individual is tempted to imagine himself detached from the world, disengaged from the common forms of life, with the language interwoven with them, which are the real a priori of any intelligent behaviour at all, never mind representing and depicting.19

Wittgenstein, like Hegel, therefore, sees the modern conception of the knowing subject as not only perpetuating an error, but as also having existential implications, as enacting an antipathy to life. More explicitly than Hegel, he speaks of metaphysics as a disease, for which 18

Rose (2009 [1981]), p. 214.

19

Kerr (1997 [1986]), pp. 134–35.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

152

Agnosticism

therapy must be administered, and a cure effected: ‘Philosophy is a battle against the bewitchment of our intelligence by means of language . . . The philosopher’s treatment of a question is like the treatment of an illness.’20 We have seen that Hegel’s re-orientation of philosophy leaves no conceptual place for agnosticism as traditionally conceived. But we have identified the ways in which a re-conceived and, in many ways, a more radical form of agnosticism or unknowing may be seen to be intrinsic to his understanding of the System. Wittgenstein’s own philosophical revolution likewise proscribes the possibility of conventional agnosticism, and for the same reasons, namely, that it is predicated on the whole modern apparatus of the abstracted knowing subject speculating on whether an object, entity, or being called God exists. Wittgenstein, as much as Hegel, considers this to be not only a mistaken epistemological model, but also a confused way of asking the question of God. But to what extent does Wittgenstein allow for, or even demand, as we saw Hegel does, a new and transformed conceptualization of agnosticism or ‘unknowing’? Is there anything like an equivalent to Hegel’s unknowing in relation to absolute knowledge in Wittgenstein’s thought? At first sight, this might appear to be an unlikely prospect. For is Wittgenstein, much more than Hegel, not thought to be a philosopher for whom ‘nothing is hidden’, ‘everything is open to view’, and all metaphysical transcendence is melted away? In other words, it might be thought there is no possibility of agnosticism or even of unknowing for Wittgenstein because ultimately there is nothing to be agnostic about. But there is perhaps more to be said here, and it will be instructive to begin with the shift from the early Wittgenstein of the Tractatus to the later Wittgenstein of the Philosophical Investigations, among other works. How is the transition from the early to the late Wittgenstein to be thematized? Slavoj Žižek has proferred an instructive conceptualization based on Jacques Lacan’s psychoanalytic thought, and in particular his ‘formulae of sexuation’. According to these formulae, there are two ways of structuring totality and difference. The masculine formula posits an exception, but it is this very exception which serves as constitutive for a unified and totalizing whole. In other words, it is the existence of the exception that constitutes, allows

20

Wittgenstein (1958), pp. 47, 91.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

153

for the possibility of, the universal. On the other hand, the feminine formula is defined by the paradox of the ‘non-All’ (pas-tout). Here, there is no exception, no founding abjection, and for that very reason, the field is non-totalized, non-unified, ‘non-All’. Žižek suggests that we should understand the shift from the early to the late Wittgenstein precisely as a shift from a masculine to a feminine structural logic in these Lacanian terms. Interestingly, for our purposes, Žižek presents this in terms of the shifting status of the Ineffable, the unknown, that about which we confess ignorance, adopt a stance of agnosticism. This is manifested in the Tractatus by the well-known and over-quoted last sentence: ‘That of which we cannot speak, we should remain silent.’ It is precisely by positing this Ineffable exception about which nothing can be said, that the unified totality of the Tractatus can be established: ‘in the early Wittgenstein of the Tractatus, the world is comprehended as a self-enclosed, limited, bounded Whole of “facts” which precisely as such presupposes an Exception: the mystical Ineffable which functions as its Limit.’21 In this sense, the mystical Ineffable serves as something like Hegel’s absolute knowledge: it both is the limit of thought and serves as its precondition. But in terms of its content, it is strikingly different; whereas Wittgenstein’s Limit evokes silence, Hegel’s manifests itself in the concrete and particular movements of consciousness. But in Wittgenstein’s later thought, of course, the mystical Ineffable is abandoned, and so too is the ‘bounded Whole of facts’ for which it is the precondition. As Žižek presents this, the loss of one necessarily entails the loss of the other, at least in terms of Lacan’s formulae: in late Wittgenstein, the problematic of the Ineffable disappears, yet for that very reason the universe is no longer comprehended as a Whole regulated by the universal conditions of language: all that remains are lateral connections between partial domains. The notion of language as a system defined by a set of universal features is replaced by the notion of language as a multitude of dispersed practices loosely interconnected by ‘family resemblances’.22

So a corollary of Žižek’s analysis is that the early Wittgenstein invoked a strict agnosticism (the mystical Ineffable) of absolute silence so that 21

Žižek (2012), p. 756.

22

Žižek (2012), p. 756.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

154

Agnosticism

agnosticism—understood as any form of ‘unknowing’—might be banished from the totalized world of facts. In contrast, the later Wittgenstein abandons agnosticism altogether, with the result that a totalized unity is lost, and we are left with a sheer multiplicity in which all is, in principle, open to view. There is no agnosticism in the form of the Ineffable, neither is there any agnosticism—anything in principle unknowable—in the manifest world. Certainly, in Wittgenstein’s later thought, much of his effort is directed against the notions of things beings uncertain, in doubt, unknowable. This is particularly the case at the beginning of On Certainty, where he is concerned with the doubts and uncertainties that seem to afflict professional philosophers. They are preoccupied with doubts about which non-philosophers are oblivious; and the ‘ordinary people’ are the ones who have got it right, and who are mercifully free of such illusions. Philosophers are explicitly in thrall to the notion of the abstracted knowing subject, and this creates all kinds of metaphysical illusions and false problems—unnecessary doubts foremost among them. These include classical philosophical doubts about what I really know, whether objects really exist, the truth of realism versus idealism, and so forth. These problems all arise from the falsely metaphysical way of viewing the world, itself arising from the dualism between the world and the knowing subject. This false view creates false problems, or at least severely misstates what might otherwise be genuine questions: is it an adequate answer to the scepticism of the idealist, or the assurances of the realist, to say that ‘There are physical objects’ is nonsense? For them after all it is not nonsense. It would, however, be an answer to say this: this assertion, or its opposite is a misfiring attempt to express what can’t be expressed like that. And that it does misfire can be shewn; but that isn’t the end of the matter. We need to realize that what presents itself to us as the first expression of a difficulty, or of its solution, may as yet not be correctly expressed at all.23

But once freed of such intellectual captivity, one will come to see that the ‘mystery’ or ‘unknowing’ which these problems express are illusions. One need only look at how a form of life is lived and how a language game is played, and one will see all that one needs to see

23

Wittgenstein (1979), p. 7.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

155

and know all that needs to be known. That anything is hidden, mysterious, unknown, or unknowable is simply an illusion. This applies equally to the question of God. Within the religious form of life itself, the question of agnosticism does not arise. For Wittgenstein, the difficulty with agnosticism is that it is a concept that has been formulated outside the religious form of life and based on a misunderstanding of what ‘God’ actually means within that form of life. Agnostics treat the existence of God as a hypothesis which, in abstraction from the form of life or language game in which it is used, may or may not be true, and finding the evidence for that hypothesis inconclusive either way, they adopt the stance of agnosticism. But for Wittgenstein, the starting point is already confused: ‘We don’t talk about hypothesis, or about high probability. Nor about knowing . . . It is for this reason that different words are used: “dogma”, “faith”.’24 The important thing is to look at how talk of belief in God operates within the religious form of life. When one does that, one sees the distinctive way in which talk of belief in God operates: If the question arises as to the existence of a god or God, it plays an entirely different role to that of the existence of any person or object I ever heard of . . . there is this extraordinary use of the word ‘believe’. One talks of believing and at the same time one doesn’t use ‘believe’ as one does ordinarily. You might say (in the normal use): ‘You only believe—oh well . . . ’ Here it is used entirely differently; on the other hand it is not used as we generally use the word ‘know’. If I even vaguely remember what I was taught about God, I might say: ‘Whatever believing in God may be, it can’t be believing in something we can test or find means of testing’.25

Within the religious form of life God is the reality or backdrop against which the whole of life is lived, not a ‘thing’ that may or may not be there, or a hypothesis that may or may not be justified. In light of this, for those who do not participate in that form of life, it is not that they deny the hypothesis that there is a thing called God, or are agnostic about it. At least, insofar as they do, Wittgenstein would say that they are confused, suffering again from metaphysical illusions. Rather, ‘I think differently, in a different way. I say different things to myself. I have different pictures.’26

24 26

Wittgenstein (1978), p. 57. Wittgenstein (1978), p. 55.

25

Wittgenstein (1978), pp. 59–60.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

156

Agnosticism

It would seem, then, that Wittgenstein rigorously extirpates all forms of agnosticism from his thought—whether the conventional form of agnosticism in relation to God or a wider philosophical ‘unknowing’. Both, in fact, are symptoms of the same disease. There is nothing hidden or unknowable; all that is to be known is laid before us. The re-conceived form of agnosticism we found in Hegel appears to be completely absent in the later Wittgenstein. Having said all of this, however, we should hesitate before acquiescing too quickly in this portrayal of Wittgenstein as a thinker of total transparency, and before drawing too sharp a contrast between Wittgenstein and Hegel in this respect. Rowan Williams has reminded us that when Wittgenstein rejects totalizing explanations of reality that claim to be all-encompassing and which claim to decode the underlying ‘truth’ of various human forms of life, he is not rejecting interpretation as such in favour of a passive reception of what presents itself but is rather defending the pluriform vitality of interpretation . . . Understanding, explaining, interpreting are not efforts of an individual to penetrate a surface: they are social proposals for common reading and common, or at least continuous, activity (a gesture or performance which in some sense goes on with or takes up from mine). They do not, therefore, see what is interpreted as setting a problem to be solved . . .27

So Williams reminds us that for Wittgenstein, we are not consigned to being simply passive participants in forms of life and passive observers of what presents itself to us. Interpretation is both necessary and enriching; but what has to be remembered is that the interpretative proposal is precisely one that is made at a point in time and space; it acknowledges the finite and so acknowledges other possible voices; it is, in fact, suspicious of a suspicion that looks for a determinate hidden content to consciousness or phenomena.28

What Wittgenstein is suspicious of is a unifying discourse that would serve to encompass and explain everything before us. What Wittgenstein is denying, in effect, is Hegel’s absolute knowledge. But given what we have already said about Hegel’s own strictures on absolute knowledge, namely, that it is not pre-judged, that it cannot be stated, that it has not been realized, we can see that even in denying Hegel’s 27

Williams (2007), p. 190.

28

Williams (2007), p. 190.

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

157

absolute knowledge, Wittgenstein is not so far apart from him, after all. Both Hegel and Wittgenstein would agree that every pretence to a universal, totalizing, and unifying account is always projected from a particular time and place, and precisely because of that, fails in its universalizing task. Furthermore, Hegel and Wittgenstein would also agree that a genuinely universalizing account is not possible for us as finite human beings. The difference is that Hegel characterizes this finite impossibility as absolute knowledge, whereas Wittgenstein refrains from speaking of it at all; presumably, for him, to do so would already be to concede too much. The result is that whereas Hegel incorporates what we have been characterizing as an agnosticism, an ‘unknowing’, into his System in the form of absolute knowledge, Wittgenstein ultimately refuses to do so.

IV What is at stake in this difference? I have been suggesting that Hegel and Wittgenstein are not as far apart as may initially be thought. Both are seeking to overcome the same misguided epistemology and metaphysics, which they both see as falsely abstracting the life of the mind from its surrounding context. Both see this epistemology as being false and existentially damaging. Both seek to effect a reunification of mind with reality, and both deny the possibility of a unified account of the totality of knowledge, at least for finite human being situated within time. In all these ways, there is much common ground, even if it is very differently expressed. But on the question of a re-conceived agnosticism—formulated by Hegel in terms of his Notion or ‘absolute knowledge’, and absent in Wittgenstein—we have registered a clear difference. Even here, as Rowan Williams’s comments remind us, the difference is not as great as might be thought. For Wittgenstein too recognizes a certain opacity arising from what Williams called the ‘pluriform vitality of interpretation’, a sense of the provisional, and the impossibility of ‘final’ or ‘ultimate’ interpretations, in spite of his insistence on transparency, and on ‘all being open to view’. Nonetheless, even if not absolute, the difference remains: Hegel’s system incorporates a ‘reconceived agnosticism’ in the form of the unrealized absolute which is both the limit and the precondition of thought, something which is completely absent in Wittgenstein’s later

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

158

Agnosticism

philosophy. I want to suggest that there is something significant at stake in this difference. Hegel’s conception of absolute knowledge implies that there is something more suggested by our forms of life and consciousness than the sheer fact of those forms of life and consciousness themselves. Those configurations are suggestive of something more. This does not imply some hidden transcendent realm or some pre-existent teleology, but something that arises from within those forms of consciousness themselves. As we look, following the quasi-revelatory ‘sunburst’, at the various configurations of thought in history, we begin to suspect that there is more than just the sheer following of one form of consciousness by another. We begin to suspect that there is a relationship between these ‘moments’, that the movement of this succession is more than just random or arbitrary. The movement itself is suggestive of something, some meaning that ‘exceeds’ each singular moment of consciousness alone. This ‘excessive’ meaning unifies the various configurations of thought, and in that sense totalizes them into a whole. But when we try to state what this ‘excessive’ meaning is, what this totalized view consists in, we immediately fail, because we immediately attempt to articulate it from our particular standpoint, and hence it becomes another form of subjective projection, another Sollen. This failure is what is signalled by absolute knowledge, which cannot be stated because it has not been realized. This failure of thought is what I take to be Hegel’s re-conceived agnosticism, a failure of thought far more radical than that found in any conventional form of agnosticism. There is no equivalent in Wittgenstein’s thought because he does not consider the question of a temporal or successional relationship between forms of consciousness or language games. He certainly considers the question of their spatial or structural relationships, drawing particular attention to their differential character, giving rise to what Žižek called a ‘multitude of dispersed practices’. But he is less concerned with the question of how and why one form of life gives way to another, and with what might be revealed in this movement. The fact that Hegel does consider this is his great merit. For it allows him to be cognizant of a fundamental feature of human experience, namely, our desire to totalize and our failure to do so. This feature of our experience can never be resolved, but neither can it be denied. This paradox has been extensively discussed by Thomas Nagel. He says that there is a human desire to aspire to an objective standpoint;

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

159

to see the world from a position that includes our own subjective selfhood within it. This standpoint is the position of what Nagel calls the ‘objective self ’: however we may have come to it, and however incomplete our development of its capacities, it places us both inside and outside the world, and offers us possibilities of transcendence which in turn create problems of reintegration. The reconciliation of these two aspects of ourselves is a primary philosophical task of human life—perhaps of any kind of intelligent life.29

But if this is so, Nagel is sceptical, to say the least, that this task can be realized. He asks: what should be the relation between the beliefs we form about the world, with their aspirations to objectivity, and the admission that the world might be different from the way we think it is, in unimaginable ways? I believe we have no satisfactory way of combining these outlooks. The objective standpoint here produces a split in the self which will not go away, and we either alternate between views or develop a form of double vision.30

It is not enough, for Nagel, simply to subordinate the objective self to the subjective one; that is to say, simply to reduce every ‘objective’ view to its ‘subjective’ projection, which would be, in effect, to reduce all to subjectivity. On the contrary, Objectivity is not content to remain a servant of the individual perspective and its values. It has a life of its own and an aspiration to transcendence that will not be quieted in response to the call to reassume our true identity . . . . The objective self is a vital part of us, and to ignore its quasi-independent operation is to be cut off from oneself as much as if one were to abandon one’s subjective individuality. There is no escape from alienation or conflict of one kind or another.31

Thus, the tension is one that cannot be resolved. There can be no solution other than acknowledging the tension and living in the midst of it: ‘Our problem has in this sense no solution, but to recognize that is to come as near as we can to living in light of the truth.’32 And this, I am suggesting, is what Hegel does. He recognizes our tendency to

29 31

Nagel (1986), p. 66. Nagel (1986), p. 221.

30

Nagel (1986), p. 88. Nagel (1986), p. 231.

32

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

160

Agnosticism

presuppose a totalized view of reality, together with our inability to state it. Hence, he points towards the totalizing vision of absolute knowledge, while insisting that it cannot be stated because it has not been realized. On the other hand, I am also suggesting that this is what Wittgenstein lacks; he rejects the notion of a totalizing perspective altogether. In light of what has been said above, is this to repress a central feature of what it is to be human? Wittgenstein rejects the notion of an independent reality which our language games and forms of life are seeking to reflect and articulate, as indeed does Hegel. But then, for Wittgenstein, reality is, for us, constituted by our language games and forms of life without reserve. Reality is for us what our language games take it to be, pure and simple. We are, or should be, completely immersed in our forms of life and any attempt to transcend them is dismissed as an illusion. Any attempt to think beyond the confines of our language games, to gesture towards the totality of consciousness is disallowed. One way of stating the difference between Hegel and Wittgenstein in this regard is to say that whereas Hegel allows for the question of such a ‘beyond’, such a ‘totality of consciousness’, but insists upon our inability to answer it, Wittgenstein simply refuses both the question and the answer. But in enacting this double refusal, an important aspect of human life and thought is lost. An answer may not be possible for finite human beings in space and time, but there may still be value in asking the question, and it may be part of our nature as human beings to do so. Indeed, for Hegel, there is something revelatory about asking this question, even without a stated or realized answer. Furthermore, Wittgenstein’s refusal in this respect leads us to ask about the status of his own voice and his own pronouncements in relation to the language games and forms of life he describes. When Wittgenstein makes his positive substantive denial of that which metaphysics asserts, and when he equally substantively asserts that the everyday, the open to view is all there is, whence is this voice articulated and what is its status? What seems to be lacking here is a proper account of its own reflexivity. What I am suggesting, therefore, is that Hegel allows for the question of ‘absolute knowledge’ to be asked (even if it cannot be answered) in a way that Wittgenstein does not. In this sense, Hegel

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

Hegel, Wittgenstein, and Agnosticism

161

recognizes a fundamental feature of what it is to be a thinking human being in a way that Wittgenstein does not sufficiently do. Furthermore, this same question of ‘absolute knowledge’ allows Hegel to be reflexive about his own voice in a way that Wittgenstein is not. Hegel is able to gesture towards this ‘absolute knowledge’ while simultaneously acknowledging that he too is situated and unable to articulate it, while the status and position of Wittgenstein’s own pronouncements remains unclear. It should be clarified that these arguments do not in any way constitute a ‘rejection’, still less a ‘refutation’ of Wittgenstein’s thought. On the contrary, Wittgenstein’s works will continue to be invaluable resources offering unique insights for those thinking through the implications of the demise of post-Cartesian metaphysics. But it is to suggest that Hegel’s attempt to undertake a similar task has some critical advantages over Wittgenstein’s own. Ultimately, his rendering of absolute knowledge as both limit and precondition, as presupposed but unrealized, is a summation of these advantages. All of which, then, suggests a continuing need for a reconceived and specifically Hegelian form of agnosticism or unknowing.

BI B LI OGR APHY Butler, Judith. Subjects of Desire: Hegelian Reflections in Twentieth-Century France (New York: Columbia University Press, 1987). Hegel, G.W.F. Phenomenology of Spirit, tr. A.V. Miller (Oxford: Oxford University Press, 1977). Hyman, Gavin. A Short History of Atheism (London: I. B. Tauris, 2010). Kerr, Fergus. Theology after Wittgenstein (London: SPCK, 1997 [1986]). Nagel, Thomas. The View from Nowhere (Oxford: Oxford University Press, 1986). Phillips, D.Z. Introducing Philosophy (Oxford: Blackwell, 1996). Rose, Gillian. Hegel contra Sociology (London: Verso, 2009 [1981]). Shanks, Andrew. Hegel and Religious Faith: Divided Brain, Atoning Spirit (London: T & T Clark, 2011). Shanks, Andrew. A Neo-Hegelian Theology: The God of Greatest Hospitality (London: Routledge, 2014). Williams, Rowan. Wrestling with Angels: Conversations in Modern Theology (London: SCM Press, 2007). Wittgenstein, Ludwig. Philosophical Investigations, tr. G.E.M. Anscombe (Oxford: Basil Blackwell, 1958).

OUP CORRECTED PROOF – FINAL, 20/8/2020, SPi

162

Agnosticism

Wittgenstein, Ludwig. Lectures and Conversations on Aesthetics, Psychology and Religious Belief, edited by Cyril Barrett (Oxford: Basil Blackwell, 1978). Wittgenstein, Ludwig. On Certainty, edited by G.E.M. Anscombe and G.H. von Wright, tr. Denis Paul and G.E.M. Anscombe (Oxford: Basil Blackwell, 1979). Žižek, Slavoj. Less than Nothing: Hegel and the Shadow of Dialectical Materialism (London: Verso, 2012).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

6 Aquinas and Agnosticism Paul O’Grady

Surprisingly little attention has been paid to clarifying the nature of agnosticism.1 As a stance in the philosophy of religion, it clearly emerged in the nineteenth century, although the earlier philosophical work of Hume and Kant paved the way for it. However, there is also an older, related tradition in philosophy and theology, called apophaticism, which makes the notion of ‘not knowing’ about God central to its concerns.2 How do these approaches relate to each other, if at all? To attempt an answer to this question I shall explore an interpretation of the work of Thomas Aquinas which emphasizes apophaticism, and a related interpretation of his work which results in one of the most systematically articulated versions of contemporary agnosticism.

I. AGNOSTICISM Contemporary agnosticism has boundaries on both sides, to theism and atheism. The boundaries are porous and so it has been challenged from both sides. Some distinctions can be made to help clarify the concept. While theism and atheism might be characterized as metaphysical or ontological claims, stating what does or does not exist, 1 A good place to survey scholarly work on agnosticism is the bibliography in Robin Le Poidevin (2010). 2 An excellent introduction to the apophatic tradition is Turner (1995).

Paul O’Grady, Aquinas and Agnosticism In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0007

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

164

Agnosticism

agnosticism is an epistemological claim, arguing that we are not in a position to make those specific metaphysical claims. As an epistemological claim, various further aspects of it can be specified. One dimension is whether it is temporary or permanent. One might lack knowledge in an area but hope or expect to acquire further information to resolve the issue, or one might think there is no possibility of a resolution. Philosophical agnosticism is typically of the latter kind, whereas the acquisition of new data characterizes scientific inquiry. Therefore agnosticism in the philosophical sense is typically concerned with conceptual or dialectical considerations, which explains why it gets short shrift from those with little regard for such matters, like Dawkins.3 Another dimension of agnosticism is its scope. To what extent is our knowledge limited? Typically one might limit one’s agnosticism to a domain—say ethics or religion while accepting knowledge claims in other domains, for example, science. How does it then differ from scepticism, which seems to have an identical profile? Robin Le Poidevin suggests that while there are similarities between the two views, they differ in agnosticism being a state of mind, whereas scepticism is a method.4 This seems to suggest that agnosticism (paradoxically) is a belief state, while scepticism is a method which leads to that. Le Poidevin says one might think of scepticism as a cause with agnosticism an effect, noting that there are other routes to or causes of agnosticism. However, there is a different way of characterizing the relationship of agnosticism to scepticism that offers an interesting way of making sense of two distinct approaches to religious belief. Agnosticism is primarily associated with lack of knowledge—as its etymology attests. Scepticism on the other hand is associated with the suspension of judgement. These two phenomena do not always go together and hence one can distinguish two significantly different kinds of agnosticism. One version leads to the suspension of judgement, which can be labelled sceptical agnosticism. The other is compatible with some positive beliefs, but it colours or modifies the way in which those beliefs are held. This can be called apophatic agnosticism. As specified so far, agnosticism has no particular connection to religious belief—it could be about ethics or mathematical ontology

3

Dawkins (2006), p. 46ff.

4

Le Poidevin (2010), p. 38.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

165

(indeed one of the main recent contributions to clarifying the concept has focused on maths).5 Yet it has been predominantly associated with religious belief. With respect to this domain, one can distinguish a number of different approaches which can be characterized as involving agnosticism. Using the distinction cited above between kinds which involve suspension of judgement and those which do not, one can distinguish two types of sceptical agnosticism and two types of religious agnosticism.

(a) Insouciant Agnosticism This is characterized by a lack of engagement with the topic and in some ways overlaps with unreflective atheism. That is, there isn’t a great deal of reflection or engagement with argument, but it reflects a more basic attitude, which may be cultural or psychological, of lack of interest in the topic: it just isn’t very interesting or important; and therefore it’s not worth devoting much time to dealing with the issues. Although actually defending robust atheism, Daniel Dennett articulates something of this attitude when he says ‘I decided some time ago that diminishing returns had set in on the arguments about God’s existence, and I doubt that any breakthroughs are in the offing, from either side’.6 It’s clear that he regards religion as something interesting (but mistaken) and needing to be combatted by scientific means, but the attitude of being dubious about the worth of the philosophical arguments is one which captures this initial constituency within agnosticism.

(b) Sceptical Agnosticism This second group differs from the first in regarding the issues as important and also in thinking that philosophical argument and analysis is relevant to them. Like theists and atheists and unlike insouciant agnostics, they place importance on the issues. One might contrast this with the attitude of many contemporary philosophers to the sub-field of philosophy of religion, which they regard as peripheral. Sceptical agnostics continue holding to the importance 5

Rosenkranz (2007), pp. 55–104.

6

Dennett (2007), p. 27.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

166

Agnosticism

of the field, but unlike their historical forebears on both sides, defend the suspension of judgement, using philosophical means. An outstanding exemplar of this approach in contemporary philosophy is Anthony Kenny.7

(c) Fideistic Agnosticism Since at least the Reformation, there is a recognizable strand of thought which seeks to limit reason to make room for faith. It is frequently associated with Protestant thinkers and can be identified in Kant and Kierkegaard.8 Drawing the limits to reason is a dominant theme in modern thought, whether through transcendental psychology or linguistic analysis. And in the absence of determining reasons one way or another there is the possibility of appealing to another resource to fix one’s beliefs, namely Revelation. Specifying the nature of reason, of faith and of their relation admits of a wide variety of positions. In recent years postmodern challenges to the grand narratives of reason have once again been recruited to support the claims of faith and flourished in such movements as Radical Orthodoxy.9

(d) Apophatic Agnosticism There is also a deep strand in mainstream religious belief which maintains that we do not have knowledge of God, that claims made about God are necessarily inadequate, that in principle no representation can capture God’s essence. While this strain of thought has a famous exemplar in Meister Eckhart, it is rooted in the thought of Pseudo-Dionysius and both Eckhart and Aquinas were influenced in Dionysian thought by Albertus Magnus.10 Historical debates have been conducted about the relative roles of Platonic and Aristotelian themes in Aquinas and about the correct relationship of cataphatic to apophatic elements in his work.11 For this present essay, I wish to explore the apophatic theme in Aquinas and its relationship to 7 8 9 10

For a clear statement of his views, see Kenny (2006), chapters 3–5. For a good study of this, see Evans (1998), chapters 5–7. See Milbank, Pickstock, and Ward (1999). 11 For a study of this, see Turner (1995). See Rocca (2004).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

167

agnosticism. It has little in common with insouciant agnosticism. Also, arguably, it is to be distinguished from fideistic agnosticism in its respect for reasoning and philosophy—although making sense of the exact relationship of faith to reason in Aquinas is a flourishing cottage industry.12 My interest in this chapter is to think about connections and differences between the sceptical agnosticism and the apophatic agnosticism especially in relation to Aquinas’ work. To do this I need to flesh out both positions in more detail. In section II I shall discuss Anthony Kenny’s defence of sceptical agnosticism. In section III, I shall look at key texts in Aquinas which point in an agnostic direction and in section IV, I shall look at some of the debates in the secondary literature about Aquinas’ putative agnosticism. What is most interesting is that Kenny’s stated grounds for his agnostic stance are exactly those which inspire Aquinas’ apophatic approach. So in section V, I want to argue that the crucial difference between Kenny’s agnosticism and Aquinas’ apophaticism is in their different conceptions of existence, a difference which reflects divergent views on technical issues about predication and reference.

II. KENNY’S SCEPTICAL AGNOSTICISM Anthony Kenny has articulated a clear, comprehensive, and detailed position with regards agnosticism across a significant number of books. Raised a Catholic, he trained as a priest in Rome and began a doctorate on religious language there. His doctoral supervisor was Bernard Lonergan, who encouraged him to engage with the texts of Aquinas, since his undergraduate training in textbook Thomism positively discouraged such a reading. He returned to England and began a second doctorate at Oxford, on the intentionality of psychological terms, while also working as a priest in Liverpool. He was influenced by Herbert McCabe and Peter Geach to read Aquinas through an analytic lens and was profoundly affected by Elizabeth Anscombe’s tutorials, where he was grounded in the work of the later Wittgenstein. 12

See, for example, O’Grady (2014), chapter 3.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

168

Agnosticism

As a priest he came to doubt that God’s existence could be proved by purely rational means and as a result didn’t take his Roman doctorate, since to do so would involve taking an oath affirming that such could be done. He was influenced by logical positivism initially and subsequently by the later Wittgenstein’s work to doubt the intelligibility of much religious discourse. His celebrated book, The Five Ways,13 analyses Aquinas’ famous arguments but finds them wanting. He argues that their context, embedded in ancient physics, makes them vulnerable to scientific advances and that they also exhibit logical flaws. Further books on divine attributes hold that the classical theistic attributes of omnipotence, omniscience, and omnibenevolence cannot consistently go together and so nothing like the God of classical theism could exist. The idea of a disembodied mind is also something Kenny thinks very hard to make intelligible. Various responses are possible to the arguments made by Kenny, but most of the responses, especially those on the intelligibility of the divine nature, rely on the doctrine of divine simplicity, which is the claim that there is no distinction between essence and existence in God. This position is one which Kenny attacks in a sustained manner. He holds Aquinas to be one of the great philosophers of the canon, but argues that his views on being are garbled and unintelligible. His Aquinas on Being14 seeks to substantiate this by a detailed historical analysis of a sequence of texts where Aquinas articulates his views on being and existence. Kenny carves up the landscape of theism, atheism, and agnosticism in an interesting way. Those who believe that ‘God exists’ is true are clearly theists. Those who do not believe this have a number of distinctions among them. There are those who positively believe that there is no God, make a distinctive metaphysical claim and so can be called positive atheists. Others who don’t accept that God exists but who don’t hold that God doesn’t exist either can be called negative atheists. One group of these thinks that question is meaningful but just have no grounds to answer it—these are agnostics. The other group think that it is meaningless and therefore is not susceptible of either true or false answers. These are positivists. Why then isn’t Kenny straightforwardly an atheist? He thinks the claim inherent in atheism is unreasonably strong, stronger than

13

Kenny (1969).

14

Kenny (2002).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

169

theism. The atheist holds that no possible definition of God could ever be made into a true existence claim. The theist claims that some specific version is true. Kenny thinks that neither of these positions have been demonstrated and so agnosticism is the most reasonable stance to take. As he puts it ‘a claim to knowledge needs to be substantiated, ignorance need only be confessed’.15 This is somewhat misleading as it seems to attribute no argumentative burden to the agnostic, a claim seemingly refuted by Kenny’s impressive list of publications. Furthermore he is critical of a number of philosophical positions which support atheism. He challenges the kind of scientism, associated with the philosophy of Quine, which goes under the label ‘naturalism’. This raises debate about the nature of philosophy. Kenny contends that philosophy is not involved in gaining new bits of knowledge, but rather in attempting to understand the knowledge we already have. It involves understanding, but understanding at a very high level of generality. This he thinks is an in-principle difference between philosophy and science. He also thinks that NeoDarwinians are over-confident in extending their methods to attempt to explain the origin of the cosmos (and indeed the origin of language). While evolutionary theory is a powerful way of making sense of living creatures, its over-extension is problematical and, indeed not obviously incompatible with the existence of something outside the natural realm. Kenny’s knowledge of the philosophical tradition of theism, his insiders-view (so to speak) of a religious form of life and his sensitivity to the appeal of a religious conception of reality make his views extremely important for a consideration of agnosticism, being a paradigm of the position I have called sceptical agnosticism. However, there is one final reason why analysis of Kenny’s views have special relevance to understanding Aquinas’ views. His carving up of the religious landscape doesn’t quite capture his own position. He is neither a theist nor an atheist but is an agnostic. However, his worries are largely about the intelligibility of the language used in respect to God. He thinks positivism is a failed project and rightly generally rejected within philosophy. Hence he says:

15

Kenny (2002) p. 21.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

170

Agnosticism

I want to bring out the difficulty of speaking coherently about God not from the hostile position of the positivists but from the traditional doctrine of the ineffability of God. The doctrine that, in some sense, it is quite impossible to speak about God: that God is not something to be captured by human language.16

If this is so, what then is the difference between the two forms of agnosticism distinguished above—Kenny’s rigorous agnosticism and Aquinas’ apophatic position—since Aquinas’ position also affirms the ineffability of God? To this I shall now turn.

III. AQUINAS’ TEXTS It seems counterintuitive to think that there might be agnosticism in Aquinas’ works. He is a saint, a doctor of the church, and most famous as someone who has written a very great deal about God. However, there is scholarly debate about this and some interpreters have emphasized a strong agnostic strain in his work. Reputable interpreters such as Brian Davies point to statement like ‘In this life our minds cannot grasp what God is: whatever way we have of thinking of God is a way of failing to understand God as God really is’17 and adds ‘The extent to which the Summa Theologiae is shot through with this conclusion cannot be overemphasized’.18 However, Eleonore Stump warns that caution is warranted about such claims. Aquinas clearly rejects a radical agnosticism associated with Moses Maimonides and makes what she calls a ‘huge, substantive, positive, metaphysical claim about God’s nature’, namely that God is a selfsubsistent form.19 How might one pull together these differing assessments, or indeed wherein lies the difference? In section III(a) shall address these interpretative debates, but first I want to assemble some texts in chronological order where Aquinas makes claims that certainly sound agnostic. There is a proliferation of relevant texts, so I am selecting certain central ones and seeking to show a continuity in theme throughout his writings. 16 18 19

17 Kenny (2004), p. 31. Aquinas, Summa Theologiae (ST) 1.13.11. Davies (2014), p. 70. See also Davies (2013). Stump (2003), p. 94, on Aquinas, ST I.3.3ad3.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

171

(a) Commentary on the Sentences In his Commentary on the Sentences (SS), written as a requirement for becoming a Master of Theology at the University of Paris (which he did in 1256), Aquinas discusses the nature of divine existence in Book 1 distinction 8. As an objection he considers whether the term ‘esse’ is appropriately predicated of God. He cites John of Damascus, who says that ‘qui est’ does not signify what God is, but rather signifies a certain infinite ocean of being. Since the infinite is not knowable, it remains unnamed and unknown and therefore is not a name of God. In response to this objection, Aquinas says that ‘qui est’ is the most appropriate of all names for God, citing Exodus, Dionysius, Avicenna, and Damascence himself. Replying to this specific point of Damascene, he notes that we proceed by ‘remotionis’ in dealing with God. First we negate all corporeal properties, then all intellectual properties such as they are in creatures, for example goodness or wisdom. All that remains is that God is and nothing more (‘quia est et nihil amplius’). He goes on to say: Finally this ipsum esse, as it is in creatures, is also removed, and so it remains in a certain darkness of ignorance, according to which, in this present life, we are best united to God, as Dionysius says, and this is a sort of thick fog in which God is said to dwell. (SS 1.8.1.1ad4)

So Aquinas claims that we have a darkness of ignorance about God and that relative to our corporeal form of life, God dwells in a thick fog. This first text highlights well a puzzle many experience in reading Aquinas on this topic. On the one hand there are the agnostic claims and terminology associated with unknowing (ignorance, fog) and on the other a clear commitment to the reality of God. Determining whether this is a consistent position is a goal of this chapter.

(b) Commentary on Boethius’s De Trinitate In Aquinas’ Commentary on Boethius’s De Trinitate (c.1257‒9) he discusses the possibility of there being a science of divine things (with the understanding that Scientia is a technical term referring to Arisotle’s conditions established in the Posterior Analytics). Against this he cites Dionysius to the effect that God is best honoured in silence. Dionysius and Damascene are major figures in Eastern apophatic theology and Aquinas quotes them extensively in dealing with

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

172

Agnosticism

knowledge or lack of knowledge of God. In response to the objection he approves of Dionysius’ claim, but says: God is indeed respected by silence, this does not mean that we may think or say nothing at all about him, but that we must realize that he always transcends anything we can think or say about him.20

Again note the two-fold move of allowing that something can be said, but simultaneously claiming that it will be in principle deficient. Aquinas adumbrates in Commentary on Boethius’s De Trinitate a three-fold approach to God, which is rooted in Dionysius—involving causation, eminence, and negation. He argues for the existence of some reality using causation, but specifies what that reality is like using negation and eminence. It is unlike creatures in straightforward ways (e.g. it is not material or changing). However it is also unlike in more subtle ways—for example, goodness and wisdom (perfections) which occur in creatures are not lacking to it, but do not exist in it in the same way in which they exist in creatures. Gregory Rocca offers a detailed analysis of the different ways this three-fold process occurs in Aquinas’ texts.21 But the fundamental move is to argue causally for some underlying reality and then to deny creaturely modes of being of it. A significant connected discussion occurs in Commentary on Boethius’s De Trinitate 6.3 where Aquinas asks whether human cognition can behold the divine form itself. Drawing on his account of human cognition he holds that there is a gap in this life in our knowing the divine essence. Crucially this also applies to information coming to us from divine revelation, since it is apprehended or grasped with human cognition which is far below the divine nature. Aquinas says: Accordingly in the present life it is absolutely impossible to know the essence of immaterial substances, not only by natural knowledge but also by revelation; for, as Dionysius says, the light of divine revelation comes to us adapted to our condition.22

This means that Aquinas closes off the third form of agnosticism discussed in part 1, where lack of philosophical knowledge makes space for faith. Clearly this is complicated since he respects divine 20 21

Aquinas, Commentary on Boethius’s De Trinitate (InBDT) 2.1 ad 6. 22 Rocca (2004), p. 49ff. Aquinas, InBDT 6.3 corp.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

173

revelation and accepts it. However, the same conditions which make philosophical knowledge of what God is like unavailable (namely our cognitive deficiencies) also operate on knowledge deriving from revelation.

(c) Summa Contra Gentiles The Summa Contra Gentiles opens with a discussion of the office of the wise man, and this primarily involves an investigation of the end and origin of the universe, which is the investigation of the highest causes. However, Summa Contra Gentiles notes that this investigation is difficult, prone to error and takes much time.23 The process of arguing from observable effects in the world to the existence of God is ‘absolutely inadequate to manifest the substance of God’.24 He goes on to say: the human reason is related to the knowledge of the truth of faith (a truth which can be most evident only to those who see the divine substance) in such a way that it can gather certain likenesses of it, which are yet not sufficient so that the truth of faith can be comprehended as being understood demonstratively or through itself. Yet it is useful for the human reason to exercise itself in such arguments, however weak they may be, provided only that there be no presumption to comprehend or demonstrate. For to be able to see something of the loftiest realities, however thin and weak the sight may be, is, as our previous remarks indicate, a cause of the greatest joy.25

The method of attempting to specify what God is like is explicitly negative. He argues that by removing elements that are not appropriate to God, one can get clearer about one’s conception. Now in considering the divine substance, we should especially make use of the method of remotion. For, by its immensity, the divine substance surpasses every form that our intellect reaches. Thus we are unable to apprehend it by knowing what it is. Yet we are able to have some knowledge of it by knowing what it is not. Furthermore we approach nearer to knowledge of God according as through our intellect we are able to remove more and more things from him.26

23 25

Aquinas, Summa Contra Gentiles (SCG) 1.4. 26 Aquinas, SCG 1.8. Aquinas, SCG 1.14.

24

Aquinas, SCG 1.8.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

174

Agnosticism

Chapters 15‒27 follow this closely. Even chapter 28 which discusses God’s perfection construes this latter as a lack of imperfection—the negation of a negation. Later in the same work he notes that by virtue of his epistemological views it would be impossible for a creature to understand God, since the means of knowledge (the intelligible species) is necessarily limited and God is necessarily unlimited. Hence: Even we are able to reach this knowledge of God, in some sense; for we know through His effects, that God is, and that He is the cause of other beings, that He is supereminent over other things and set apart from all. And this is the ultimate and most perfect limit of our knowledge in this life, as Dionysius says in Mystical Theology. ‘We are united with God as the Unknown.’ Indeed, this is the situation, for, while we know of God what He is not, what He is remains quite unknown. Hence, to manifest his ignorance of this sublime knowledge, it is said of Moses that ‘he went to the dark cloud wherein God was’. (Exod. 20:21)27

A number of distinct issues emerges in these quotations. Some knowledge of God’s existence is possible, but knowledge of God’s nature is not available through human reasoning. It seems that faith does allow deeper knowledge, but not knowledge that is demonstrative. And it seems that there may be a distinction between knowledge in this life and in the next, ‘And this is the ultimate and most perfect limit of our knowledge in this life.’ Hence in making sense of Aquinas’ claims on human knowledge of God, the issue of degrees of knowledge and the context of that knowledge (pre/post mortem) will be relevant.

(d) Summa Theologiae The Summa Theologiae (1265‒73) is regarded as the mature articulation of Aquinas’ views. One of the clearest statements of Aquinas’ views is to be found in the prologue to question 3 of the Prima Pars. There he says: Once it is known that something exists, it remains to be investigated how it exists, in order that what it is may be known. But because it is not

27

Aquinas, SCG 3.49.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

175

possible for us to know what God is, but rather what God is not, we cannot consider how God exists, but rather how God does not exist.28

In a manner similar to that of Summa Contra Gentiles, the subsequent discussion of God’s nature follows the via remotionis, with a clear statement of a fundamental lack of composition in God, the lack of a distinction between God’s act of existing and God’s nature.29 In the opening question on the nature of Sacra Doctrina, Aquinas states that ‘we cannot know what God is’ (non possimus scire quid est)30 and later says ‘through the revelation of grace we do not in this life now what God is, and thus we are joined to him as to one unknown’.31

IV. STATING AQUINAS’ POSITION Sufficient texts have been cited to show that there is indeed a clear and consistent position in Aquinas’ works claiming that we do not have knowledge of God’s nature. And this is not merely a contingent lack of knowledge, but has to do in principle with the disparity between our cognitive abilities and what God is like. And the lack of knowing also applies equally to whatever knowledge gleaned by revelation as well as to whatever philosophical knowledge is available. However, this is deeply puzzling. Aquinas clearly articulates views about God’s attributes, as well as the doctrine of divine simplicity which underpins his whole view of God. He also subscribes to the Incarnation, the Trinity, and indeed has a very strong requirement that one explicitly and verbally endorse the articles of the creed. So how is that at all reconcilable with the texts cited about not knowing God’s nature? Aquinas explicitly distinguishes his view from those of Moses Maimonides, who expressed a stronger negative approach to God. Aquinas rejects the claim of Moses, that no positive terms whatsoever can be predicated of God, arguing that we do distinguish some terms as more appropriate of God than others (e.g. good versus hairy).32

28 30 32

29 Aquinas, Summa Theologiae (ST) 1.3 prol. Aquinas, ST 1.3.4. 31 Aquinas, ST 1.1.7ad1. Aquinas, ST 1.12.13. Aquinas, ST 1.13.2.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

176

Agnosticism

The basis of all claims Aquinas makes is causal. He thinks that creation has a causal relationship with God. On this basis he argues for the lack of potentiality in God, hence the deduction of pure actuality, perfection, immutability, eternity, etc. Yet these can be construed as forms of negation. Nevertheless they seem clear knowledge claims, so how does this fit with the language of not knowing? In the background are technicalities from Aristotle’s account of scientific knowledge (Scientia). Standard Scientia proceeds from knowledge of causes to effects. Demonstratio propter quid, full scientific demonstration, is a full explanation of effects through their relationship to their causes. This is also known as quidditative knowledge—knowledge of the deep inherent structures of reality. Given this is an idealization and may only exist in purely formal sciences such as logic or geometry, it nevertheless stands as the gold standard for knowledge. Aquinas denies that this kind of knowledge of God is available to us. We know about God from effects. Very general observable phenomena such as motion, change, generation, grades of being, and order require some kind of explanation which is of a different kind to themselves. This is demonstratio quia, a form of knowledge which is much less than full scientific knowledge. On the basis of causal arguments for this deeper level of explanation, the features highlighted in the effects allow one to make more precise the underlying explanatory cause—it is immaterial, unchanging, etc. However, these features do not yield Scientia in the manner of demonstratio propter quid. So Aquinas can deny that knowledge (understood as scientific quidditative knowledge of the deep nature of things) of God is possible. Hence it looks as if Aquinas is saying that some knowledge of God is possible at a level which is below that of proper Scientia. There is the affirmation of the existence of some mysterious reality which is the result of causal arguments and the specification of some features of that reality which are largely negative (i.e. it is not x,y,z). He also allows for the possibility that greater knowledge may be possible after death. Human happiness, he believes, is tied up with knowledge of God and if this is not possible, then ultimate happiness is not achievable. Furthermore there is the inbuilt desire to know which would be frustrated if God were in principle unknowable. Therefore God is capable of being seen by the blessed, but even then they cannot comprehend him—there is no exhaustive knowledge possible. And

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

177

this post-mortem knowledge is not available on purely natural grounds, but requires the lumen gloriae.33 Returning to the issue of knowledge of God in this life, it is evident that Aquinas affirms that there is some knowledge of God but also some lack of knowledge. Timothy McDermott captures this by drawing a nice contrast. He says that agnostics typically have a worked out view of what God is like but are not sure whether there is such a thing. On the other hand, Aquinas is sure there is something mysterious at the basis of reality, but is in principle unsure about what it is like.34 A certain tradition of interpreters have focused on these negative themes in Aquinas, have connected it to considerations about language, and used certain tools from analytical philosophy to articulate their insights. At the head of this tradition one can find Victor White, an English Dominican theologian with unusual interests. He corresponded with C.G. Jung, interested himself in Freudian and Jungian psychology, and wrote a book called God and the Unconscious (1952). He also wrote a very influential book called God the Unknown (1956) where he anticipates McDermott’s view saying: St Thomas’s position differs from that of modern agnostics because while modern agnosticism says simply, ‘we do not know and the universe is a mysterious riddle’, a Thomist says ‘we do not know what the answer is, but we do know that there is a mystery behind it all which we do not know, and if there were not, there would not even be a riddle. This Unknown we call God. If there were no God, there would be no universe to be mysterious, and nobody to be mystified.35

White was an influential teacher, the Regent of Studies at Blackfriars Oxford and there went on to influence several of his successors in that role, including Herbert McCabe, Fergus Kerr, and Brian Davies. While White himself didn’t deploy analytical philosophical methods, each of these others did. And this way of bringing together Aquinas and analytical philosophy also had a profound impact on Anthony Kenny, who says that at Oxford he ‘met the Dominican Herbert McCabe and the Birmingham philosopher Peter Geach, who soon convinced me that it was possible to combine the techniques of linguistic analysis with an appreciation of the teaching of St Thomas’.36 Once again the question arises, how does Kenny’s 33 35

Aquinas, ST 1.12.1–4. White (1956), p. 18.

34 36

McDermott (2011). Kenny (2004), p. 4.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

178

Agnosticism

position differ from that of these interpreters of Aquinas who emphasize apophaticism and the use of linguistic methods? I shall attempt an answer to this in section V.

V. ISSUES ABOUT EXISTENCE While Kenny consistently champions Aquinas as a great philosopher, he also thinks that Aquinas’ views on being and existence are confused and unworkable.37 Since Aquinas’ views on God are closely linked to his views on being, Kenny also believes Aquinas’ natural theology is flawed. What are the problems with Aquinas’ position? Kenny espouses Frege’s analysis of existence and believes that it is properly understood via the mechanism of quantification in language.38 There has been some debate as to whether it is appropriate to use Frege’s work in this way and that it constitutes a kind of hermeneutical violence to Aquinas’ work to interpret it in such a different idiom.39 I agree with those commentators who argue that Kenny clearly thinks Frege is right and it is not simply a matter of looking at different interpretations of texts.40 Kenny thinks Aquinas’ alternative views are mistaken when seen in the light of Frege’s analysis. So one needs to engage with Kenny’s specific arguments against Aquinas on existence. Kenny distinguishes twelve different senses of being, or of the verb ‘to be’, in Aquinas’ work:41 1) is Specific existence, which asserts a kind or a type of being; 2) is Individual existence, referring to a particular individual, for example ‘The Great Pyramid still exists’; 3) is Substantial being, where some individual is identified as a substance; 4) is Accidental being, where the verb ‘to be’ attributes something from the other Aristotelian categories than substance; 5) is Common being, which is what absolutely everything which exists has in common, a very thin and universal predicate; 6) is Actual being, where something potentially the case is brought to realization; 7) is Absolute being, the kind of being unique to God; 8) is Intentional being, which is the mode of being something has while in the mind; 37 40

Kenny (2002), p. v. Pugh (2006), p. 268.

38 41

Kenny (2002), p. 195ff. Kenny (2002), p. 189ff.

39

See Klima (2004).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

179

9) is Fictional being, which pertains to privations (blindness) and also to logical kinds such a genera and species; 10) is Possible being— Kenny is doubtful whether Aquinas held this type of being, but some early passage might allow for it; 11) is Predicative being, which deals with the verb ‘to be’ as a copula, joining and predicate to a subject; and 12) is Identical being—which involves a reversibility in ‘A is B’ not seen in other uses of the verb. His overall complaint is that Aquinas doesn’t offer a coherent synthesis of how all these fit together. However, what is relevant in this chapter is that in dealing with claims about God’s existence, Kenny claims there are several plain errors in Aquinas’ treatment. Probably the most fundamental confusion he identifies is that between Specific and Individual existence. He thinks that Aquinas (and various important Aquinas scholars such as Owens and Wippel) confuse these two kinds of being in thinking about the famous claim that in God there is no distinction between essence and existence.42 Is the type of existence mentioned in ‘there is no distinction in God between essence and existence’ to be understood as either Specific existence or Individual existence? If it is Specific existence, the thought is that in this unique case, the kind of thing in question just exists, with no other specification involved. For Kenny this is an absurdity—it results in an ill-formed formula (∃x) x, the use of the existential quantifier without any predicate to attach to it. So the claim makes no sense on that construal. On the other hand if one treats it as Individual existence, it seems to suggest that the existence of this individual is included in its essence. But this claim takes us back to Anselm’s argument, where an atheist may cheerfully accept that classical theists hold the view that God’s essence includes its existence, but just deny that any such entity exists in reality. Furthermore, if God’s existence is understood as Common existence, that which absolutely everything has in common, then it emerges as the thinnest, most atrophied, least interesting property possible. If understood as Absolute existence it takes us back to the distinctive existence associated with God’s essence—which is purely conceptual and compatible with there being no God. So Kenny thinks there is no construal of ‘God’s essence is existence’ which makes sense. In fact he

42

Kenny (2002), p. 43 fn26.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

180

Agnosticism

thinks Aquinas flip-flops between different construals such that his view is characterized by ambiguity and equivocation. There have been a number of responses to Kenny, arguing that his criticisms lack purchase. One can note that Aquinas is very careful to distinguish between individuals and kinds when dealing with Platonism. He consistently denies that kinds exist outside the mind (although the real pattern or form which constitutes the kind does exist outside the mind, but not as a separate abstract entity). Concrete individuals are the prime instance of being in Aquinas’ system. So it is a bit odd to accuse him of confusing something he consistently distinguishes. All the realm of genera, species, etc., have their existence in Intentional being (in the mind which conceives them) and so are ontologically dependent on that individual mind. The real patterns conceptually represented by that mind depend on the concrete individuals in which they are instantiated. So there seems no incoherence so far. For Aquinas, God is an individual, albeit one whose metaphysical structure is uniquely different to anything else. The kind of existence in God is not possessed by God, but is God, whereas the kind of existence possessed by anything else is distinct from the kind of thing it is. So it seems appropriate to characterize God’s existence as Individual and also Absolute. Kenny’s objection was that this is a merely conceptual point, God’s existence is analytically contained in his essence but that doesn’t lead to God existing. Well, first, Aquinas emphasizes that we do not know God’s essence, hence we can’t deduce things from it (thus blocking the ontological argument). His arguments for God’s existence start with observation and argue back to a primary cause. Thus the existence of God is rationally the result of a dialectical argument about explanation, not the result of conceptual analysis. However, the core difference between Kenny’s construal and Aquinas’ is that Kenny has a bipolar understanding of existence, whereas Aquinas has a scalar understanding. On the bipolar view, one either exists or not at all; no third option is available. Hence the way existence applies to every single existent is the same in each instance—the thinnest possible predicate. Aquinas, on the contrary, thinks of existence in terms of actuality. Each thing’s existence is different in virtue of the degree of actualization present in it. How might one understand this? Each concrete individual has distinctive structure, captured in the idea of essence. But this is understood as

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

181

potential relative to the act of existence which makes it real. That essence and existence are really distinct doesn’t mean that they must be numerically separable entities. Rather, the claim is that they are really distinct metaphysical parts of any given individual entity which when united yield an existent entity. Neither are to be understood as a property, but rather both can be understood as principles of being (taking principle to mean basic explanatory element). Essence is related to existence as potency to act. And each act of existence differs from another by virtue of the essence which it actualizes. This allows for conceiving existence in scalar terms, the greater the actualization, the greater the existence. Matthew Pugh argues that the doctrine of the transcendentals is relevant here in making sense of the scale of being.43 The transcendentals are alike in extension, but differ in intension. For example, goodness and being refer to the same things, but the aspect picked is different. Goodness refers to the desirability of things, existence to the actuality of things. But the thought is that the perfections that are the positive properties which a thing has are contained within its actualized being. So on this conception, esse is not the thinnest possible property, but actually the most profound, a claim Aquinas makes in the De Potentia, ‘esse est quod est profundius’.44 Kenny thinks that Aquinas’ account of existence is not coherent and therefore cannot serve to articulate a view on the nature of God. Defenders of Aquinas argue that it is and that the crucial difference is between a bipolar versus scalar conception of existence. For Kenny, since no coherent position emerges, the only legitimate option is to suspend judgement and be a sceptical agnostic. Those who emphasize the apophatic dimension of Aquinas’ work, argue that his account of existence is sufficiently rich to allow for the postulation of the existence of an ultimate reality whose existence is largely unknown by virtue of its superiority to our conceptual capacities. Yet what of the paradoxicality in holding to the apophatic approach while simultaneously affirming all sorts of positive views about God? Is Aquinas in some sense or other intellectually split, by turns apophatic and positive?

43

Pugh (2006), p. 272ff.

44

Aquinas, De Potentia 7.2ad9.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

182

Agnosticism VI. THE PURPOSE OF THE APOPHATIC DIMENSION

There are at least four ways in which the apophatic dimension in Aquinas’ work does useful conceptual work. The first is in addressing a familiar problem in contemporary philosophy of religion, the relationship between the God of the philosophers and the God of religious faith. Pascal articulated this in a sharp form. Philosophy results in bloodless abstraction, whereas real religious experience is a visceral, burning phenomenon, gripping one’s entire being.45 However, such a dichotomy is alien to Aquinas. There is a continuity between his religious practice and his philosophical speculation. Indeed he refers to the emotional intensity of intellectual speculation about God, ‘For to be able to see something of the loftiest realities, however thin and weak the sight may be, is, as our previous remarks indicate, a cause of the greatest joy’ (Summa Contra Gentiles 1.8). As a Dominican, his form of life involved a complex structure of prayer and practice awash with scripture and indeed some more exotic forms of medieval piety such as self-flagellation and copious tears, which is anything but a bloodless abstraction. I believe the apophatic dimension in his work allows for the interaction of these parts. His philosophical theology is not an enclosed rationalistic system where one can ascend to the deity by algebra, as it were. At its core is mystery, ineffability, and indeed silence. There is therefore the possibility of connecting this to the positive, scripturally-based, religious tradition he inhabited precisely because it is not self-enclosed and finished. This quality of openness derives from the apophaticism. Second, this dichotomy between the God of the philosophers and the God of religion has been re-interpreted in the post-Heideggerian critique of ontotheology. The basic objection there is that when Greek metaphysics is brought to bear on questions about God, God is subsumed under a wider framework, in which God exists as a part. God becomes a kind of deistic explanatory category and removed of mystery. The apophatic strand in Aquinas’ thought makes this an unsustainable criticism, despite the fact that the rationalistic systematization of Aquinas’ views makes it initially plausible. Jean-Luc Marion, a leading post-Heideggerian phenomenologist initially 45

See Pascal (1995), p. 178.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

183

characterized Aquinas’ work as ontotheological, but withdrew this claim on attending to the apophatic dimension of his thought.46 Third, the work of the medieval German Dominican, Meister Eckhart (1265‒1327), has in recent years received much attention in terms of its mysticism, its influence on later German philosophy (Boehme, Hegel, Schopenhauer, Heidegger) and its potential connections with Buddhism. It is a rich field of study albeit not much covered in the analytical tradition. Eckhart expressed daring and paradoxical statements such as ‘I pray God to rid me of God’ and ‘God is a pure nothing’. While accused of heresy in the early fourteenth century, recent interpreters have sought to argue that Eckhart’s work can be seen as part of the Dionysian tradition and given orthodox meaning. The two sayings just cited can be glossed as saying that conceptual thought is incapable of capturing the true nature of God and that God is not to be understood on the metaphysical model of thinghood. Understood in this way, Eckhart’s views are completely congruent with Aquinas’. And indeed there is a strong historical connection between their work.47 Eckhart frequently cites Aquinas, was strongly influenced by Aquinas’ teacher and Dionysian scholar, Albertus Magnus (1200‒80), and very unusually held the Dominican chair of theology at Paris twice, the only other figure prior to him to do that being Aquinas. The apophatic dimension in Aquinas therefore makes his work more interesting from the perspective of history of mysticism and for interreligious dialogue, particularly with nontheistic religious traditions. Finally, Eleonore Stump has recently suggested a useful way of making sense of how the positive and negative dimensions in Aquinas’ work relate to each other. She appeals to the famous waveparticle duality model which is familiar in modern physics.48 In making sense of the behaviour of light at a sub-atomic level, it is acceptable for physicists to describe it sometimes as having characteristics of a wave, sometimes having characteristics of a particle. These are not regarded as being contradictory, but as illuminating aspects of an underlying reality which is not fully understood. Stump suggests that the positive approach to God emphasizes the personal aspects of the deity, while the apophatic approach emphasizes the

46 48

Marion (2012), chapter 8. Stump (2012), p. 139.

47

Turner (1995), p. 143.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

184

Agnosticism

impersonal, ineffable dimensions. Her claim is that both work as mutual correctives of each other and can be held in creative tension.

VII. CONCLUSION Agnosticism is an interesting phenomenon and admits of a variety of interpretations spanning a spectrum of religious and irreligious attitudes. By distinguishing different kinds and examining the worked-out examples of Aquinas and Kenny one can see the kinds of considerations that are relevant to taking a stand on whether and how to have religious belief. Technical considerations about meaning, about how to construe existence and about analyticity have an impact on whether one can intellectually commit to belief in a mysterious reality which is held to be of fundamental importance, but which is in principle unknowable. As well as being a challenge to insouciant agnostics, Aquinas’ views are also a challenge to those theists who are perhaps overly sure of their grasp of the deity, many of whom swell the ranks of contemporary philosophy of religion.

BI B LI OGR APHY Davies, Brian. ‘Aquinas and Atheism’. In The Oxford Handbook of Atheism, edited by Stephen Bullvant and Michael Ruse (Oxford: Oxford University Press, 2013), pp 119–38. Davies, Brian. Thomas Aquinas’s Summa Theologiae: A Guide and Commentary (Oxford: Oxford University Press, 2014). Dawkins, Richard. The God Delusion (London: Bantam Press, 2006). Dennett, Daniel. Breaking the Spell: Religion as a Natural Phenomenon (London: Penguin, 2007). Evans, Charles Stephen. Faith beyond Reason (Edinburgh: Edinburgh University Press, 1998). Kenny, Anthony. The Five Ways: St. Thomas Aquinas’ Proofs of God’s Existence (London: Routledge, 1969). Kenny, Anthony. Aquinas on Being (Oxford: Oxford University Press, 2002). Kenny, Anthony. The Unknown God (London: Continuum, 2004). Kenny, Anthony. What I Believe (London: Continuum, 2006). Klima, Gyula. ‘Kenny on Aquinas on Being’. International Philosophical Quarterly 44, no. 4 (2004), pp. 567–80.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Aquinas and Agnosticism

185

Le Poidevin, Robin. Agnosticism: A Very Short Introduction (Oxford: Oxford University Press, 2010). Marion, Jean Luc. God Without Being (2nd edn) (Chicago: University of Chicago Press, 2012). McDermott, Timothy. ‘Review of Mark Vernon How to be an Agnostic’. Times Literary Supplement, 16 September (2011). Milbank, John; Pickstock, Catherine, and Graham Ward (eds.). Radical Orthodoxy: A New Theology (London: Routledge 1999). O’Grady, Paul. Aquinas’s Philosophy of Religion (London: Palgrave, 2014). Pascal, Blaise. Pensées, edited by Anthony Levi (Oxford: Oxford University Press, 1995). Pugh, Matthew. ‘Kenny on Being in Aquinas’. In Analytical Thomism: Traditions in Dialogue, edited by Craig Paterson and Matthew Pugh (Aldershot: Ashgate, 2006). Rocca, Gregory. Speaking the Incomprehensible God (Washington, DC: Catholic University of America Press, 2004). Rosenkranz, Sven. ‘Agnosticism as a Third Stance’, Mind 116, no. 461 (2007), pp. 55–104. Stump, Eleonore. Aquinas (London: Routledge, 2003). Stump, Eleonore. ‘God’s Simplicity’. In The Oxford Handbook of Aquinas, edited by Brian Davies and Eleonore Stump (Oxford: Oxford University Press, 2012). Turner, Denys. The Darkness of God, Negativity in Christian Mysticism (Cambridge: Cambridge University Press, 1995). White, Victor. God the Unknown (London: Chatto and Windhus, 1956).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

7 Redeeming Agnosticism: Richard Kearney’s Anatheistic Wager Guy Collins

I. INTRODUCTION Unlike its more certain counterparts, theism and atheism, agnosticism is plagued by an intrinsic difficulty: how deep does not-knowing go? Or to put it another way, if theism and atheism are structurally alike in being discourses of knowledge, each describing a position in reference to God, are there any limits to the uncertainty of the agnostic? While theists and atheists may argue with one another about what it is that they actually believe in, and whether they are correct in that belief, in the popular imagination at least, agnostics seem to have the opposite problem. They appear not really to believe in anything. Which is, ironically enough, as was observed long ago, much the same problem as believing in pretty much anything. If theism and atheism are strategies of knowing, which rule some beliefs in and some beliefs out, the problem with agnosticism seems to be deciding when it can rule on anything, whether in or out. By contrast, as Richard Kearney consistently argues across many different books, but most notably in Anatheism: Returning to God after God, it is possible to break away from the idea that theism and atheism are mutually exclusive poles of knowing. Anatheism builds on deep uncertainties and limits to the knowledge claims of theism and atheism in order to present an alternative strategy of notknowing. This chapter argues that such a strategy of not-knowing Guy Collins, Redeeming Agnosticism: Richard Kearney’s Anatheistic Wager In: Agnosticism: Explorations in Philosophy and Religious Thought. Edited by: Francis Fallon and Gavin Hyman, Oxford University Press (2020). © OUP. DOI: 10.1093/oso/9780198859123.003.0008

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

187

can nevertheless continue to offer profound possibilities for religious and ethical transformation. The paradox of agnosticism is that agnostics have an important perspective on matters of theism and atheism. At the same time, the question remains whether agnosticism can ever be something other than just a failure to choose a position between theism and atheism. In what follows, I present a thought-project inspired by the specific philosophical-cum-theological project of Richard Kearney’s anatheistic wager. The central question of this thought-project is whether it is possible to identify a form of agnosticism that is redeeming. Specifically, I will be exploring whether agnosticism can be theologically generative by assisting the human quest for God. In contrast to views of agnosticism that treat it as an exclusively psychological phenomenon, I shall proceed on the assumption that agnosticism can help us better understand the metaphysical and epistemological aspects of atheism and theism. Unlike many other philosophers and theologians who self-identify as theists or atheists, Kearney chooses instead the neologism of anatheism to describe his personal and intellectual beliefs. While Kearney generally eschews the language of agnosticism, in developing the question of anatheism Kearney raises important questions about the centrality of the intersection between doubt and belief. Faced with doubts about the divine, anatheism offers a way of weaving the experience of unknowing (agnosis) into deepening our understanding of the divine. My hope is that through this examination of Kearney’s anatheism it may be possible to challenge conventional understandings of agnosticism as a barrier to making sense of the divine. There are several major components to Kearney’s anatheistic argument, and he brings in a wide variety of philosophical, literary, religious, and theological voices into play. One important idea animating this chapter is whether anatheism can be read as a new type of agnosticism. Specifically, we will explore whether Kearney’s anatheism goes beyond a simple failure to choose between theism and atheism. I hope to show that Kearney’s anatheism is more than an inability to take a position. Nor is it a position of studied neutrality that takes neither theism nor atheism seriously. Instead, I argue that Kearney’s anatheism redeems the unknowing of agnosticism by showing it to be theologically generative. By investigating anatheism through the lens of agnosticism I contend that Kearney’s project can revitalize and expand our understanding of both agnosticism and

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

188

Agnosticism

theology. The capacity to ‘not know’ and to be uncertain is essential when thinking about the divine. Moreover, anatheism offers a possibility of re-thinking through the increasingly sterile distinction that sees theism and atheism as self-contained, and mutually exclusive, discourses of knowledge. As will be seen, one of the central intellectual resources that powers Kearney’s anatheistic argument is a diverse set of voices from within continental philosophy. Alongside his study of classic literature and sacred scriptures, without continental philosophy it would be impossible to make the anatheistic argument. Agnosticism, atheism, and theism are all capable of generating religious meaning. However, this is particularly evident when there has been an encounter with a continental philosophy that reprises the invitation of the mystics to discover a God without being. Meister Eckhart’s encouragement to let go of God in order to find God finds its contemporary refrain in the continental philosophical approaches to the divine that underlie Kearney’s anatheistic project. Much work has been done exploring the connections between negative theology, apophasis, and continental thinkers. Kevin Hart, in particular, trailblazed by exploring remarkable similarities and overlaps between Derridean deconstruction and the apophatic tradition of negative theology.1 One of Hart’s unique insights was to view negative theology as a type of deconstruction. For our current purposes, what is most pertinent about Hart’s analysis is how apparently radically different discourses may be connected at a deeper level. An analogous question about potential connection hovers over the rest of this chapter: specifically, whether or not there is a resemblance between agnosticism’s unknowing and apophatic not-knowing shared by both mystics and continental philosophers. The agnostic way, the mystical way, and the atheistic way are usually thought of as separate ways. However, mysticism and atheism are both tasked with rejecting certain models of God and as a result of this overlapping activity have sometimes been confused. As a result, mystics like Meister Eckhart have sometimes been charged with atheism, while a figure like Jacques Derrida has been the subject of much fruitful religious analysis despite professing that ‘I quite rightly pass for an atheist’.2 Part of the work of this chapter is to explore whether an anatheistic version of agnosticism might offer a better way of

1

Hart (2000).

2

Derrida (1993).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

189

approaching the inherent overlaps between ‘religious’ figures such as Eckhart and ‘atheistic’ figures like Jacques Derrida. The two have very different approaches to belief. And yet, as Hart revealed, in denying the ability to know the divine they have common interests. My question is whether such radically different denials of knowing the divine (mystical and atheistic) cannot imply a connection to an alternative, agnostic, way of limiting divine knowing. To back up a little, Kearney’s anatheism conserves and affirms both the Eckhartian and the Derridean. Anatheism includes the witness of both those who have religious experiences of the divine and atheistic deniers of the possibility of any such experience. This ability to traverse both atheism and theism is of profound importance to understanding anatheism. And yet, despite the originality of Kearney’s broad-minded anatheistic embrace of atheism and theism, Kearney does not disclose a new theological position. To be absolutely clear: Kearney is not creating a tertium quid here in anatheism that is neither theistic nor atheistic. Instead, Kearney is reframing, and creatively drawing together, theological and philosophical writings that simultaneously bifurcate both atheism and theism. That these are both atheistic and theistic reveals that what is truly known here remains open. Anatheism is not a strategy for knowing one particular thing in relationship to theology or the divine. Instead, it is a reminder that religious knowledge cannot be certain, and that it is imperative to keep one’s options open. If anatheism is a practice of double hospitality to atheism and theism alike, might not this also be the source of its striking resemblance to agnosticism? Unlike either theism or atheism there are no obvious philosophical standards or hermeneutical communities, theological or otherwise, charged with overseeing the practices and beliefs of agnostics. Part of the point of being agnostic, one assumes, is that unlike theist and atheist the agnostic is not ready or willing to take a specific stand on questions of metaphysics. At the same time, it is not hard to imagine an agnostic who is unsure about the existence of a particular kind of God, just as it is fairly easy to imagine an agnostic who is unsure about the non-existence of any kind of God. In other words, it seems perfectly reasonable for an agnostic to believe all sorts of things about the existence or non-existence of God.3 There are many different 3 Le Poidevin’s (2010) overview of agnosticism focuses primarily on introducing different approaches to philosophical approaches to agnosticism. On the theological

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

190

Agnosticism

types of agnostic, just as there are many different types of theist and atheist. But what makes an agnostic an agnostic rather than an unsure theist or an unsure atheist? I wish to suggest that this question is in itself unhelpful, and that there may not be much of a difference. To suggest that agnosticism refers to a third position that mediates between the certainties of theism and atheism seems to me to miss the point. Instead, I contend that far from being a separate belief system, agnosticism may be viewed as a phase within both theism and atheism that opens them up to one another in fruitful dialogue. Indeed, from the perspective of Christian theology, agnosticism is a necessary phase within Christian belief. In choosing phase as the metaphor with which to describe the interrelationship of atheism and theism I do not mean it in the sense that fashion goes through certain historical phases that quickly fade away as they are superseded. Instead, I have in mind the scientific meaning of phase as a relationship in time between something that oscillates or repeats, such as an alternating electrical current. It is not my intent to suggest that agnosticism is a phase that can be passed through to a higher state, whether that is theism or atheism. This would be to mistake agnosticism for a type of Hegelian movement of sublation or aufgehoben that overtakes all that is previous to it, drawing it into a ‘higher’ realm. However, it may well be that agnosticism can be likened to a continual phase of oscillation between theism and atheism that presents some essential truth about the divine. Kearney’s Anatheism reveals what this oscillating movement connecting atheism and theism might look like. As Kearney states it: ‘My wager throughout this volume is that it is only if one concedes that one knows virtually nothing about God that one can begin to recover the presence of holiness in the flesh of ordinary existence’.4 The striking thing about this statement is that it is both an utterly agnostic statement about the limits of knowledge of God, and also a remarkably theological statement about building on this lack of knowledge to discover ‘holiness’. In this manner, anatheism casts a new light on agnosticism as a way of approaching the divine that is simultaneously deeply sceptical about what can be known about God while side, Weatherhead (1965) and Pattison’s (1996) provide different ways of theologically engaging with unknowing. 4 Kearney (2011), p. 5.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

191

passionately committed to incorporating the limits of that knowledge as the place where the holy can once more be incorporated into ‘the flesh of ordinary existence’. What makes Kearney’s anatheism potentially redeeming is that it does not flee engagement in the world as the site where we encounter the holy so much as recognize the importance of a sacramental dimension intrinsic to agnosticism. While for Kearney the identity of the holy is undecidable it also repeatedly breaks through in the material world, in a sacramentality that reveals the transcendent in the immanent. While other versions of agnosticism or atheism are constitutionally incapable of recognizing such sacramental moments, the fact that they are an essential part of anatheism makes it intriguingly theological. Further, I think it can be fruitful to conceive of atheism and theism in Venn diagram form as two circles that overlap one another. Although most of each respective circle does not overlap the other circle, the portion that does overlap is the anatheistic one. And, in turn, one can imagine another circle added to the other two that can be described as the agnostic circle. While most agnosticism is not anatheistic, anatheism is the name given to that fertile area where all the circles overlap. Richard Kearney self-describes as a philosopher rather than a theologian. However, the thought of this Irish-born Roman Catholic educated thinker is traversed by the most serious and deepest of theological categories of thought. A student of Paul Ricoeur in France, Kearney draws attention at the beginning of Anatheism to the fundamental question Ricoeur asked his students:‘d’où parlez-vous?’ (‘from where do you speak?’).5 Kearney answers this question by highlighting both his Irish Roman Catholic heritage alongside the Protestant intellectual traditions that he learned from Ricoeur and others. As Kearney concisely frames it, he is intellectually Protestant and emotionally Catholic.6 But, as is clear from the autobiographical opening to Anatheism, this is no straightforward dualism of heart versus mind. Despite having witnessed some of the worst excesses of religious certitude and division at first hand in Ireland and Northern Ireland, Kearney affirms a respect for the intellectual habits present across the Christian traditions. The Benedictine monks of Glenstal Abbey introduced Kearney to some of the most penetrating critiques

5

Kearney (2011), p. ix.

6

Kearney (2011), p. xiv.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

192

Agnosticism

of religion, those of Feuerbach, Nietzsche, Sartre, and Russell. As he recalls, before the monks approached the question of what God might be like, they first made sure in their Christian doctrine class to have students wrestle with arguments against God’s existence.7 It is significant that Kearney begins Anatheism with a review of his own presuppositions, intellectual genealogy, and religious commitments. Despite making no claim to be a working theologian, and for all that he prefers the language of philosophy, the fundamental contribution of Anatheism is to offer a new way of thinking through theological issues. While Kearney has been described as the greatest living Irish philosopher, Anatheism is altogether too theological to be a work of pure philosophy. As his deeply autobiographical preface makes clear, Kearney has drunk deeply from theological wells. Just as important, however, is how keen Kearney is to engage with theological traditions beyond the Christianity of his upbringing. Drawing on Judaism, Islam, and other religious traditions, Kearney’s anatheistic thesis engages wholeheartedly with these different faith traditions. Building on previous work, a central anatheistic theme is how each religion approaches the encounter with the stranger, and specifically the ‘uninvited guest’ as revealing the divine. While agnosticism deals with an un-knowing about God, in Kearney’s hands one of the repeated questions has to do with the others we do not know. The strangers, the unknown others, are also where religious traditions seek to locate and encounter with the holy. While Kearney engages with different theistic traditions, he is also alert to the importance of atheism as a critique of many of the most disreputable and dangerous traits within religion. Anatheism is a return to God, but it is a return to a God that arises out of a wholehearted engagement with atheism. It is worth quoting Kearney’s wager of faith in full, an important term that we will return to later: I propose the possibility of a third way beyond the extremes of dogmatic theism and militant atheism: those polar opposites of certainty that have maimed so many minds and souls in our history. This third option, this wager of faith beyond faith, I call anatheism. Ana-theos, God after God.8

7

Kearney (2011), p. xii.

8

Kearney (2011), p. 3.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

193

Unlike many partisans of the debates between theism and atheism, anatheism is Kearney’s way of opening up a new space within which both theism and atheism can sustain a more fruitful dialogue. To be clear, this third option is not an entirely new thing, as the Venn diagram suggests. Instead, it is the creative space where theism, atheism, and agnosticism already naturally intersect and overlap. The neologism anatheism describes a repetition forward of theism—a theism that has been through atheism. Ana- is the prefix that is critical here. As Kearney notes in the collection of essays dedicated to responding to Anatheism, ana- has the dictionary definition meaning of ‘Up in space or time; back again, anew’.9 Kearney’s anatheistic project thinks through what kind of God might be possible after the God of metaphysical sovereignty has been rejected, or at least found wanting in key respects. At the same time, Kearney’s openness to atheism as a fundamental movement within the anatheistic wager means that this is not a simple philosophical redescription of God. While it is not the etymology Kearney presents, it is also entirely possible to misread anatheism as ‘an atheism’ rather than ‘ana theism’. The possibility of this alternative reading indicates something of the uncertainty surrounding anatheism. Furthermore, Kearney is clear that anatheism is not a position so much as a wager. Wagers are orientations to the future, not settled positions. The tentativeness of anatheism is part and parcel of its identity, echoing the possibility (as opposed to positionality) with which God is viewed in Kearney’s The God Who May Be.10 In this sense, anatheism describes an oscillating movement that connects some of the most traditional insights of theism with some of the deepest held beliefs of atheism. Kearney cannot be certain about the anatheistic God. Instead, the anatheistic wager incorporates a deep-seated unknowing about God. Anatheism has several structures that govern the way it weaves the warp of atheism with the woof of theism. The first, as has already been noted, is that Kearney describes it as having the character of both a philosophical and an existential wager. Kearney clarifies that he has in mind the wagers of Ricoeur and Taylor rather than a Pascalian one: ‘in that they are more about imagination and hospitality than calculation and blind leaps. They solicit fidelity not fideism’.11 The wagers 9 11

10 Kearney and Zimmerman (2015), p. 6. Kearney (2001). Kearney (2011), p. xvii (emphasis in the original).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

194

Agnosticism

here are both about what can be known about the nature of God (which is very little), and also about the way in which we respond to God (which is much more). Kearney clarifies that the anatheist wager has five ‘main components: imagination, humor, commitment, discernment, and hospitality’.12 Kearney also identifies three ‘basic elements’ within anatheism itself: protest, prophecy, and sacrament.13 It is noteworthy that most of these are verbs, describing active ways in which the anatheistic wager is brought into being. Anatheism is more of a mode of orientation than a belief system. Although Kearney does wrestle at length with traditional questions such as how there can be a good God in the face of evil, it is not the identity of the deity that forms the heart of anatheism. Instead, the anatheistic project focuses most closely on the hermeneutical, philosophical, and existential realities through which individuals traverse both theism and atheism in the quest for God. Unlike much of the abstract, conceptual, and deeply theoretical thought to have come out of theology and philosophy, anatheism embodies Kearney’s insight that we never have a God’s eye view of things.14 The impossibility of ever having an objective view of God is the philosophical and hermeneutical starting point, and it informs the second part of his wager, the existential way in which humans find themselves wanting to respond to the unknowable God. What makes anatheism more than simply a reiteration of the impossibility of knowing anything about God is Kearney’s view that the God after God still bears returning to. God in some form or another keeps breaking back into texts and lives: even texts and lives apparently trying to keep God at arm’s length. Kearney has a strong sense that even the most atheistic discourse cannot simply avoid the God conversation. And, as we shall see, Kearney holds up some of the most atheistic writers and thinkers as prime examples of how the question of God keeps returning. For Kearney, anatheism is a moment that is ‘available to anyone who experiences instants of deep disorientation, doubt, or dread’.15 Human not-knowing is the entry-point to seeking the divine, not the capacity to know. In this anatheism differs from other approaches to theism and atheism. However, it is this intellectual and hermeneutical humility about the limits of knowledge about God that create deeper connections 12 14

Kearney (2011), p. 40. Kearney (2011), p. xv.

13 15

Kearney (2011), p. 85. Kearney (2011), p. 5.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

195

between anatheism and many of the world’s religions at their most prayerful, reflective, and socially transformative.

II. THE SOURCES OF ANATHEISM Before examining more closely the contours of anatheism it is important to review in broad outline the key sources. Kearney draws on diverse texts, both atheist and theist, philosophical and theological, and, perhaps most importantly, both ‘sacred’ and scriptural as well as ‘secular’ and literary. As he writes, ‘I am as likely to draw from the writings of agnostic thinkers and novelists as to invoke religious scholars and experts’.16 The diversity and variety of writers Kearney engages with reveal the truth of this statement. In the introduction alone, Kearney draws on poets and writers Oscar Wilde, W.H. Auden, Dostoyevsky, Gerald Manley Hopkins, and Hölderlin, alongside Kierkegaard and Heidegger. In subsequent chapters, Kearney ranges widely not simply in the fields of literature and continental philosophy, but also across Jewish and Muslim thinkers, as well as individuals renowned for their ‘sacramental actions’ like Jean Vanier, Ghandi, and Dorothy Day. As Kearney’s wager is developed, it is clear how anatheism offers nothing new, so much as a way of re-discovering and returning afresh to hidden riches within both atheism and theism. Underlining the importance of not-knowing in the journey of faith, Kearney explains: ‘No one is exempted from the moment of not-knowing. Anatheism presupposes this a-theistic moment as antidote to dogmatic theism. True faith, as Dostoyevsky put it, “bursts forth from the crucible of doubt.”’17 Two of the most significant oscillations at the heart of anatheism involve movements that are internal to continental philosophers and literary agnostics. On the philosophy side, Kearney engages at length with the thought of Jacques Derrida, Julia Kristeva, Emmanuel Levinas, Maurice Merleau-Ponty, and Paul Ricoeur. In the realm of literature, Kearney gives sensitive and creative readings of James Joyce, Marcel Proust, and Virginia Woolf. In what follows we will focus on the role played by continental philosophers, focusing in 16

Kearney (2011), p. xvi.

17

Kearney (2011), p. 6.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

196

Agnosticism

particular on Kearney’s examination of the Jewish philosopher Emmanuel Levinas. However, given the importance of literature to Kearney’s project, we will also examine his treatment of writer Virginia Woolf. Despite their very different beliefs and concerns, Kearney draws on both Levinas and Woolf to reveal how anatheism connects apparent opposites: both as a movement from atheism to theism, and as a movement from theism to atheism.

III. THE JEWISH INFLUENCE OF EMMANUEL LEVINAS Before analysing Levinas it is helpful to note that several Jewish thinkers are critical to Kearney’s argument. Asking the question of how it can still be possible to speak of God after the Holocaust, Kearney engages with the thought of Hannah Arendt, Jacques Derrida, Emmanuel Levinas, Elie Wiesel, Etty Hillesum, and Rabbi Irving Greenberg. Before grappling in more detail with Levinas and Derrida, Kearney first builds on different, but interrelated, insights gained from Arendt, Hillesum, and Greenberg in their struggle to account for God following the horrors of the death camps. Unable to relate to a God of sovereign power, Hillesum introduces us to a God who is found in powerlessness. Meanwhile Arendt reminds us of the need for Aristotelian phronesis, a theme of practical wisdom that is critical to Kearney’s wider work.18 Summarizing Greenberg, Kearney sees not a God who is a remote sovereign, but instead a God who is involved: Suffering with his people. From an anatheist perspective, the covenant is to be understood as a divinity calling humans to full partnership, to cocreation, or, as the old Talmudic adage had it, to the completion of the seventh day of Creation.19

18

Aristotelian phronesis is a recurrent theme in Kearney’s work. It is especially significant to the development of Kearney’s thinking in (Kearney [2002]) and (Kearney [2003]). 19 Kearney (2011), p. 61.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

197

Each of these Jewish interlocutors help Kearney frame his rejection of the God of classic philosophical theism, while opening up the possibility of another God in its wake. Together they provide a different way of exposing the traditional theistic God of omnipotence and sovereignty as a mirage. In the face of the horrors of the Holocaust, how can one ever believe in an intervening all-powerful God who is good? Kearney, like his interlocutors, does not think such a God worthy of belief. Perhaps his strongest argument here is when he quotes Elie Wiesel announcing the death knell of ‘conventional theism’, before going on to agree that the idea of God having a Divine Plan in the light of Dachau, Sorbibor, and Treblinka is a ‘cruel sham’.20 Turning to Levinas, Kearney is able to make aspects of his argument even clearer. Levinas directly names the importance of an atheism at the heart of religion. At the same time, Levinas also provides Kearney with one of the most philosophically robust rationales for the importance of the stranger (the other). For Levinas the soul itself is ‘naturally atheist’, because the soul and God cannot simply be idolatrously fused together. Levinas argues that to be a self is to be separate from God, and that to recognize the difference between God and oneself is fundamental to a proper religious understanding. We need not to be joined together with God (or to one another) precisely in order to be able to recognize the difference that is both the stranger who is other, and the holy one who is Other. We need this ‘moment of atheistic separateness’ in order to know the other as divine or as a stranger. Kearney concludes, ‘It is in this context that Levinas holds that the gift of Judaism to humanity is atheism—namely, separation from God so as to encounter the other as absolutely other’.21 There are many affinities between the thought of Levinas and Derrida. But in Kearney’s reading of Derrida and otherness, Derrida threatens to evacuate any specificity in talking about the ‘holy other’ into the utter undefinable otherness of the ‘wholly other’. Kearney criticizes Derrida for presenting a ‘mystical atheism’ that is ‘so devoid of any kind of concrete faith in a person or presence (human or divine) that it loses any claim to historical reality’.22 Levinas offers something more hopeful. For while Levinas accepts an atheistic 20 22

Kearney (2011), p. 58. Kearney (2011), p. 65.

21

Kearney (2011), p. 62.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

198

Agnosticism

moment, the absence of the sovereign omnipotent God, Levinas also offers a way of concretely encountering the holy in the world. For Kearney, the difference between Derrida and Levinas is not that Levinas is any less critical of the traditional theistic model of God as sovereign and omnipotent. After all, Levinas embraces an atheistic moment within God. But, unlike Derrida, Levinas locates an ethical commitment in the ‘face’ of the other, a specific other who is often defined as the widow, the orphan, and the stranger. For Levinas these faces of real strangers are traces of God, whereas for Derrida the question of who the other is remains entirely more nebulous.23 While Derrida is unable to determine what kind of other the other is, Levinas identifies the face of the suffering others with the face of God. This ethical commitment to strangers and others through concrete encounter with their faces informs Kearney’s vision of anatheistic hospitality. Like Derrida, Levinas demands a departure from the old theistic omnipotent sovereign God, affirming an atheistic moment within theology. Yet, unlike Derrida, this Levinasian atheistic movement leads into a concrete responsibility to the other, the site of a return to God. Levinas’s bifurcated relationship to God, of first leaving and then returning, perfectly anticipates and encapsulates the oscillating movement away from and back to God that lies at the heart of anatheism. Levinasian responsibility to the stranger runs through much of Richard Kearney’s other writing, and is particularly present in Strangers, Gods and Monsters. There Kearney anticipates the anatheistic urge to use philosophy in order to be ‘more hospitable to strangers, gods and monsters without succumbing to mystique or madness’.24 Of course, responsibility, justice and hospitality are also themes treated at great length within Derrida which Kearney explores in some detail.25 But Levinas’s view of the stranger in the face of the other is a more concrete form of otherness than what Kearney describes as Derrida’s more ‘mystical atheism’. Kearney remains critical of Derrida: ‘regardless of Derrida’s profound debt to his mentor Levinas, his purely formal messianicity prevents him from embracing Levinas’s ethical commitment to the “face” of the other— the widow, the orphan, and the stranger—as the trace of God’.26 And yet, while there is a difference between Derrida and Levinas regarding 23 25

Kearney (2011), p. 65. Kearney (2003), p. 68ff.

24

Kearney (2003), p. 18. Kearney (2011), p. 65.

26

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

199

the question of the stranger, there is no doubt that both draw attention to the oscillation between atheism and faith that is at the heart of the anatheistic wager. While Levinas offers readings of otherness more capable of being grafted onto Kearney’s Aristotelian instincts for a practical wisdom that serves the stranger, it is Derrida in Sauf le Nom that Kearney quotes as underlining the inescapable connection between atheism and theism: ‘The most consequent forms of declared atheism will always have testified to the most intense desire for God’.27 Here Derrida anticipates one part of the oscillation of the anatheistic thesis: the idea that thought that has rejected God will always have been connected to a desire for God. The fact that this thought that rejects God will always lead back again to a desire for God is, of course, simply the other phase or repetition of anatheistic oscillation. Both Derrida and Levinas have therefore equipped Kearney with significant philosophical resources to make the anatheistic movement from faith to atheism and back to faith in some kind of deity. But it is Levinas more than Derrida who ensures that the resulting agnostic view of a deity beyond atheism is also a deity located in the stranger. Philosophically, this Levinasian strain within Kearney’s thought ensures that a proper agnosticism about the nature of God does not descend into an agnosticism about how to achieve practical wisdom (phronesis) in responding to the ethical demands of the stranger. Without this Levinasian influence anatheism would run the risk of being so agnostic about the nature of the other as not to be able to know to serve the stranger in the faces of others. In his later discussion with Catherine Keller, Kearney makes it clear that he is still critical of the limits to the thought of Levinas (and Derrida) when it comes to building ‘a politics of practice’ around actually doing good, justice, and hospitality. Here Kearney prefers Keller’s idea of ‘entanglement’ (similar to his own ‘hospitality’) with its deeper relationality and reciprocity than the duty to the stranger of Levinas that has no reciprocity. But what is perhaps most fascinating about Kearney’s exchange with Keller is how it serves to highlight both thinkers’ commitment to mystics like the author of The Cloud of Unknowing and Nicholas of Cusa. Ironically, while Kearney typifies Derrida as offering a ‘mystical atheism’ that is clearly derogatory, the

27

Kearney (2011), p. 63.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

200

Agnosticism

genuine mystics of the church are held up as examples to be followed. These mystics tie us to a ‘politics of community’ that supplements their mystical agnosticism with a genuine practice of faithful community lived in the world.28 At the heart of Kearney’s distinction between ‘Derridean’ and ‘Christian’ mystical experience is not agnosis, but the question of how the religious community continues to engage in certain sacramental and communal practices even in the face of unknowing. Unlike Derrida, the author of The Cloud is therefore able to make a distinction between the impossibility of knowing God through one’s intellect and the possibility of knowing God through love. The intellectual unknowing of The Cloud does not serve to distance us from God. Instead, it serves to point out how the soul can unite with the divine and also connect with a greater community: Unite yourself to him by love and trust, and by that union you will be joined both to him and to all who like yourself are united by love to him . . . with our lady, St Mary, who, full of grace, perfectly heeded every passing moment; with all the angels in heaven, who have never let time pass; and with all the saints in heaven and on earth, who by their love, and by Jesus’ grace, take proper account of every moment.29

Both Derrida and the The Cloud affirm unknowing in relation to our ability to know God intellectually. But the mystic is able to supplement this intellectual unknowing with another experiential route into the divine that is also an experience of community. Returning to Anatheism, Kearney highlights connections between the atheism of Levinas and other Jewish thinkers and the rejection of God witnessed in the Christian mystical tradition. Kearney essentially repeats the insights of faithful rejectors of idolatrous models of God within religious traditions. While it may sound anachronistic to claim the author of The Cloud as an agnostic, the apophaticism present in The Cloud also presents a form of unknowing that goes deep. At one point we are told, ‘He [God] cannot be comprehended by our intellect or any man’s—or any angel’s for that matter’.30 Kearney’s anatheism builds on this theological apophaticism within Christianity as well as the Jewish atheism represented so well by Levinas. Both Judaism and the Christian mystical tradition are equally afraid of representing 28 30

Kearney and Zimmerman (2015), p. 65ff. Anonymous (1961), p. 55.

29

Anonymous (1961), pp. 56–7.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

201

God. The genius of anatheism is that it makes it possible to start to reframe the agnosticism of secular modernity as naturally connected to the more obviously theologically charged apophatic traditions. There is no great gulf between secular and sacred unknowing of God. Instead, apophatic theology and modern agnosis alike rest on both the possibility of atheism and the possibility that God can be known.

IV. THE AGNOSTIC INFLUENCE OF VIRGINIA WOOLF While Levinas rejects a philosophical God in order to turn back to God through the faces of strangers, widows, and orphans, Virginia Woolf presents a personal rejecting of God that offers a potential return to God through the literary figures of her texts. Once again, it is the dynamic of the movement away from God and back to God that interests Kearney as anatheistic. To the Lighthouse, Kearney suggests, performs a passage from theism to atheism—the famous interlude of death and dereliction—before reopening, in the third part of the novel, an anatheist option: the possibility of a second yes to the ‘real’ at the heart of nonbeing.31

Kearney argues that while a certain model of God has been decisively rejected, Woolf (along with Proust and Joyce) is ‘haunted by a singularly mystical vision of things’.32 This haunting manifests itself in sacramental interpretations of the world that none of the three writers can keep out of their texts. In particular, Kearney reads Woolf as suggesting that the world is not simply ‘made’ either by human or divine creators. Instead, there is an ‘intimation of some deep unfathomable love’ that undergirds the true pattern of reality.33 The anatheistic movement within Woolf is therefore one that repeats, or anticipates, the movement for Levinas: a move away from the traditional theistic God of the philosophers towards a more uncertain picture of the divine that appears within the creation. This is an anatheism that brings back the sacred in the midst of the secular: 31 33

Kearney (2011), p. 119. Kearney (2011), p. 123.

32

Kearney (2011), p. 128.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

202

Agnosticism

one that has fled philosophically loaded concepts of God, only to rediscover that which is inescapably mystical in the secular. It is significant that Kearney discerns in both Woolf ’s agnosticism and Levinas’s Judaism the anatheistic movements that challenge the simple polarity represented by traditional theism or traditional atheism. Each suggests a fresh possibility, that the sacred still exerts a call even after one has rejected the more certain and specific manifestations of both theism and atheism. In this, of course, they are not entirely alone. While there is not space to explore all the thinkers across the different traditions Kearney engages with, it is important to examine two of Kearney’s chief twentieth-century anatheistic interlocutors within the Christian tradition: his own mentor, the philosopher Paul Ricouer, and from the perspective of theology, the German pastor Dietrich Bonhoeffer. By looking more closely at how Ricouer and Bonhoeffer shape Kearney’s anatheism we can consider how it is possible to be simultaneously agnostic and faithful.

V. THE CHRISTIAN INFLUENCE OF DIETRICH BONHOEFFER Bonhoeffer’s religionless Christianity has a significant influence on anatheism. Just as Jewish thinkers rejected much of the philosophical model of a theistic sovereign and omnipotent God, so Bonhoeffer does the same for Christian theology. In introducing talk of a God ‘without religion’ Bonhoeffer started a dialogue within the Christian tradition that took issue with the God of metaphysics. The God that counts for Bonhoeffer is the suffering God, just as much as it is for Hillesum and Greenberg. This focus on the suffering of God is an approach that has been central within Christianity, across the spectrum of both time and tradition. Kearney underlines how contemporary theologians, particularly John Caputo and Catherine Keller, have gone on to develop Bonhoeffer’s original thesis of the weakness of God.34 But, of course, there are many earlier twentieth-century theologians indebted to Bonhoeffer. Chief among them must surely be Jürgen Moltmann and his exploration of the centrality of the cross 34

Kearney (2011), p. 68.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

203

and how it reveals a suffering God.35 Quoting from Bonhoeffer’s prison writings, Moltmann affirms, ‘Only the suffering God can help’.36 Part of what is interesting about Kearney’s return to Bonhoeffer is that here again he invokes the autobiographical. It is not simply Bonhoeffer the theologian who interests him. He also wants to pay homage to the embodied nature of Bonhoeffer’s religionless faith. Bonhoeffer was known by fellow prisoners as someone ‘who called for the “polyphony of life” for joy as well as sadness . . . ’.37 By foregrounding Bonhoeffer’s own lived faith, and his deep trust in the presence of God, Kearney reminds us that religionless Christianity is not a faith without God. Instead, it is faith in a God beyond the God of metaphysical certainty, a God found in powerlessness, suffering, and the plight of the stranger. Or, as Bonhoeffer puts it a ‘church for others’.38 By moving the focus of Christian belief from the God ‘out there’ of metaphysics to the God found in the sufferings of the powerless, whether prisoners or strangers, Bonhoeffer enacts something very similar to both Hillesum and Levinas.

VI. EUCHARISTIC AGNOSIS Ricoeur continues Bonhoeffer’s focus on freeing religion from theistic idolatry, exploring the philosophical possibilities of an atheism that can ‘emancipate religion from itself ’.39 By engaging with the classic hermeneutics of suspicion of Marx, Freud, and Nietzsche, Ricouer sought to show that while the God of ontotheology dies, this is not necessarily the God, to coin a classic Kearney concept, who may be.40 Critically, though, Ricoeur recognized that belief in God after encountering atheistic objections could not be the same. One cannot simply return to a traditional theism. This is why Kearney’s language of anatheism emerges as a working description for that alternative possibility Ricoeur sees emerging beyond atheism. Kearney thinks

35 36 37 39

Moltmann (1974). Bonhoeffer (1971), p. 360f. Quoted in Moltmann (1974), p. 47. 38 Kearney (2011), p. 70. Kearney (2011), p. 71. 40 Kearney (2011), p. 72. Kearney (2001).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

204

Agnosticism

Ricoeur’s argument here is ‘deeply anatheistic’, providing a route to a post-religious God unshackled by metaphysics.41 Once again, the autobiographical plays a pivotal role, as Kearney examines Ricouer’s final testament, Vivant jusqu’ à la mort. Written as Ricoeur was dying it describes ‘the grace of a certain kind of dying’, in which it was possible to feel an ‘intimate transcendence of the Essential which rips through the veils of confessional religious codes’.42 In the conclusion to his testament Ricoeur reflects on the Last Supper in an immensely important passage that deserves to be quoted in full. The Last Supper conjoins the moment of dying unto oneself and serving the other in the sharing of food and wine which joins the dying person to the multitude of survivors reunited in community. This is why it is remarkable that Jesus never theorized about this or never said who he was. Maybe he didn’t know, for he lived the Eucharistic gesture, bridged the gap between the imminence of death and the community beyond.43

This Eucharistic reflection does not highlight sacrifice as the central theological knowing of the Eucharist. Instead it presents Eucharist as the site of a deep theological agnosticism within the tradition. This Eucharistic agnosticism that Jesus opens up points instead to the lived practice of Eucharist, and the way that Eucharist draws life out of death, connecting the dying to the living, and the experience of living to that of dying. This sacramental focus is an essential part of Kearney’s anatheism, and it is noteworthy that in Ricoeur’s telling Eucharist is both profoundly redemptive and deeply agnostic. As a religious coda, it is not irrelevant to note a parallel with the sixteenth-century theologian Richard Hooker’s profession of agnosticism towards the sacramental presence of Christ in the Eucharist. Like Ricoeur and Kearney, Hooker affirms the possibility of sacramental encounter with the divine. However, he also anticipates the notion that the precise nature of how God is made present in Eucharist cannot be known. In marked contrast to Catholic, Lutheran, or Zwinglian orthodoxies of the day, Hooker laid claim to the radical idea that ‘it can be no

41 42 43

Kearney (2011), p. 75ff. Ricoeur (2007), pp. 45 and 47; quoted in Kearney (2011), p. 77. Ricoeur (2007); quoted in Kearney (2011), p. 91.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

205

disgrace to confesse wee are ignorant’.44 Hooker’s Eucharistic agnosticism is not identical to Ricoeur’s, but both sought not to undermine Eucharist with their unknowing so much as to underline the significance of participation in the sacrament over intellectual or theological interpretations of it. Hooker’s concern was not to demythologize or strip God out of Eucharist. Instead, he affirms a limited form of agnosticism of human knowing when seeking to make sense of the sacramental Eucharistic presence of God. Here we see Hooker anticipating the anatheistic sacramental space at the centre of the Venn diagram where epistemological agnosticism intersects with metaphysical theism. Some of the most significant criticisms of anatheism concern the precise content of this sacramental Eucharistic reality. At one point Kearney interprets Andrei Rublev’s icon of the three angels seated around what he describes as an empty chalice. Jens Zimmerman’s thoughtful critique notes that this chalice should instead be interpreted as a Eucharistic symbol that ‘contains the very sacrificial offering that makes hospitality essential to the Christian faith’.45 Both Zimmerman and Merold Westphal suggest that Kearney rejects the Eucharist’s sacrificial reality for something altogether emptier and, like the Derridean reading of Plato’s khora, lacking in clear definition, being neither one thing nor the other, with no clear meaning and no clear identity. Kearney, in turn, disagrees, arguing that anatheism is an attempt to underline the Abrahamic reality of divine hospitality to strangers, epitomized in Rublev’s icon. As he reflects, ‘On the contrary, the chalice of nourishment is central to my entire reading of anatheist hospitality’.46 These different perspectives on the nature of Eucharistic sacrifice draw into focus an important aspect of anatheistic sacramentality. Zimmerman and Westphal are not wrong to claim that Kearney emphasizes the indeterminacy of the Eucharist. Zimmerman remarks that ‘Kearney wants us to travel lightly, and it seems to me that perhaps he wants us to travel too lightly. And the first thing it seems to me he wants us unnecessarily to give up is the sovereignty of God’.47 At the philosophical heart of this disagreement is the question of the influence of continental philosophy, and whether, as 44 46 47

45 Hooker (1977), p. 342. Kearney and Zimmerman (2015), p. 238. Kearney and Zimmerman (2015), p. 252. Kearney and Zimmerman (2015), p. 229.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

206

Agnosticism

Zimmerman suggests, it has been responsible for creating a God that is ‘too small’ and ‘too much domesticated by continental philosophy’ rather than scripture and tradition.48 By contrast, I am suggesting that it is precisely the resources of continental philosophical that enable Kearney to shape anatheism as a type of agnosticism that remains singularly capable of remaining open to the Eucharistic reality of the Christian tradition in particular. Far from selling out on the sacraments because of the undue influence of continental philosophy, it is continental philosophy that enables Kearney to highlight and conserve the innate agnosticism of a Eucharistic hospitality. Moreover, it is a Eucharistic hospitality that by Kearney’s own account overflows the chalice: a hospitality experienced not solely in one experience of the sacrament but also in literature, other religions, and daily life. Yet, while the Eucharist represents the plenitude of God’s presence, it is not altogether surprising that the anatheist might also interpret the chalice as empty, as the impossibility of ultimately locating God is always part of the Eucharistic mystery. If Kearney did not read Ricoeur’s last testament as a deeply anatheistic and deeply Eucharistic hospitality to a life beyond death, then Zimmerman’s critique would have more force. But as Kearney shows throughout, he is concerned more with the theme of divine hospitality in the sacrament rather than with parsing any particular theological formulation of how God is present in Eucharist. All the while, Kearney retains a deep agnosticism about the God who makes this hospitality and presence possible. Zimmerman correctly notes that Kearney does not want to theorize the specifics of how Eucharist is sacrifice. However, as the example of Richard Hooker should make clear, there are Eucharistic theologies that are agnostic about the way in which God is present in the sacrament. Such theologies refrain from looking too deep into the mystery of Eucharist, while also holding on to Eucharist as the preeminent site of encounter with the holy. To put it another way, one could argue that Eucharist is a good example of anatheism avant la lettre. Eucharist both resists theological attempts to describe how God is present and, at the same time, demands that theologians see God’s hospitality in the sacrament. In this sense, the anatheistic paradox of Eucharist is

48

Kearney and Zimmerman (2015), p. 235.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

207

always how both the absence of God and the presence of God are always signified by the sacrament.

VII. INDIVIDUAL AGNOSIS One other, equally paradoxical, aspect of the sacramental within anatheism deserves attention. Anatheism is not a methodology for withdrawing the divine from life. Instead, it is a modality for experiencing the divine as a flux that sees God continually disappearing and appearing, both hiding and being revealed, suggestive both of atheistic absence and of religious encounter. And as Kearney argued effectively in Strangers, Gods and Monsters the question of the other is never simply a question about others we encounter. The other is also found within. Or to put it in the language of our current concern, there is also an internal agnosticism within each of us. We do not know who we are. We are not transparent to ourselves. For Kearney, discovering who we are is closely connected with the sacramental work of learning both how others embody the divine for us and how we can start to embody the divine for others. One of the clearest possible markers that anatheism is more than simply an agnosticism of indecision is seen again and again in the effort Kearney takes to identify those embodied moments. Kearney goes to great lengths to show how those embodied moments are to be found all over the place: not just in continental philosophy or in the different global faiths but also in works of literature, poetry, and art. And yet in addition to these examples of high culture, the central anatheist concern is altogether much more ordinary: the anatheistic return to the Eucharist of the everyday. It involves sharing bread with others and aliens surrounding us and (as Julia Kristeva reminds us) within us. For if others are strangers to us we are equally strangers to others and to ourselves.49

What ultimately characterizes Kearney’s anatheism is that it is at its core, through and through, a practical, sacramental, lived experience of how to make sense of the call of the other—in all the different forms that the other presents to us. Sounding downright Levinasian, 49

Kearney and Zimmerman (2015), p. 153.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

208

Agnosticism

while affirming the ability of anatheism to be found almost anywhere, Kearney reiterates how anatheism ‘begins and ends with the epiphany of the divine in the face of the stranger’.50 Whereas agnosticism has often been taken to be synonymous with indecision, anatheism’s oscillations and uncertainties are always rooted in this sacramental positioning that elides the question of the stranger with the question of the divine. In his response to Anatheism, John Caputo encourages viewing undecidability as ‘the condition of possibility of a deeper affirmation’ rather than as mere agnosticism.51 Just as theism and atheism can be remarkably imprecise and unhelpful terms, Caputo sees agnosticism as ‘just more jockeying, positioning, more positionality’.52 While this is certainly true of some agnosticism, possibly much agnosticism, I wish to suggest in the final section (section VIII) that Caputo’s criticism opens up a way of differentiating anatheism from the agnosticism first coined by Thomas Huxley in 1869 to describe metaphysical uncertainty.

VIII. REDEEMING AGNOSTICISM: THE ANATHEISTIC DIFFERENCE Clearly there is a world of difference between Huxley and Kearney. If agnosticism only signifies failing to take a position, an affirmation of not being able to know at all about matters of the divine, then anatheism is clearly not agnostic enough. But if agnosticism can be seen as a way of responding to questions of the divine differently to either traditional theism or atheism, then clearly anatheism is somewhere on the agnostic spectrum. Or to put it in Caputo’s terms, maybe the problem with traditional agnosticism of Huxley is not that it is too uncertain, but that is not uncertain enough. By contrast, anatheism represents a deeper, richer, and more textured agnosis: a way of approaching faith that has taken on the continental philosophical insight that the divine is, indeed, truly undecidable. 50 51 52

Kearney and Zimmerman (2015), p. 149. Kearney and Zimmerman (2015), p. 211ff. Kearney and Zimmerman (2015), p. 212.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

209

One movement of anatheism that is not always present within agnosticism is the action of returning to a God after atheism and theism. God is found within conditions of possibility that are strictly speaking undecidable, but anatheism also directs us to locate God in a hospitality towards the stranger. It seems that one reading of anatheism could be that it supplements the classic agnosticism of Huxley (that we do not and cannot know God) with the theological insight that God is indeed found, in an undecidable way, in pondering the limits of theism and atheism. As Kearney notes, this is not simply a second enchantment with the divine: anatheism is a continual moving towards and away from and back again to the divine. In this sense, anatheism is clearly different from Huxley’s agnosticism that claims nothing can be known. But neither is it reducible to classic theism or classic atheism. No reading of Anatheism can but come away with a strong sense of the author’s passion for encountering the divine by way of hospitality to the stranger. For all that post-atheistic writers think they have left God behind, Kearney notes that transcendence breaks through. Anatheism is a recognition that whether we like it or not, the sacred will not simply disappear. God returns. And God returns most especially where God appears to have disappeared, in suffering and powerlessness. The anatheist is agnostic in that it is never possible to be certain about the divine. But this does not stop the anatheist from simultaneously recognizing that in sacramental hospitality we are redeemed. God is made present sacramentally, because God is discovered not in philosophical exploration but incarnate in the everyday, disclosed through encounters with others. This disclosing of God even happens when we don’t always know what is happening. As Kearney points out in his description of Mary’s anatheistic ‘pondering’, faith is not a question of certainty. For ultimately, when it comes to faith, Kearney thinks that we are all anatheists one way or another.53 A more conventional, albeit less thought-provoking, way of putting this would be to note that we are all agnostics who have been redeemed not by our own knowledge, but by the possibility that God continues to encounter us (and we, God) through others. This would also seem to be remarkably congruent with Christian tradition. The genius of anatheism is that through close readings of religious

53

Kearney and Zimmerman (2015), p. 210.

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

210

Agnosticism

traditions, continental philosophy, and other secular sources we re-discover a God who may already be waiting for us. Recognizing that we do not know this God, and that strangers will be the ones to present God to us, is as essential to Christian theology as it is to anatheism. In rejecting both the sovereignty of the omnipotent God of classic theism and the sense that there is nothing beyond the atheist critique, anatheism strikes out an altogether new path. What matters here is how undecidability is a condition of possibility that allows engagement with both theism and atheism. While accepting both the logic and the reality of atheism, Kearney also affirms what has often been named as the very religious possibility of the impossible: specifically, the possibility that God might be encountered through the stranger and the sacraments of everyday life. This ‘possibility of the impossible’ is a classic theological trope, but one that continental philosophy has helped breathe new life into. Kearney remains aware that this possibility (of the impossible moments in which transcendence breaks through) is strictly undecidable. For him God is possibility above all. And with possibility to be there is also the possibility not to be. If anatheism came down hard on one side of the possibility of God rather than the other it would not be entirely unjust to view it alongside other theisms or atheisms. But as there is an undecidability at its heart that seeks to return to the possibility of God after both theism and atheism it seems prudent to view anatheism as instead a re-imagined experience of agnosticism. It is not the agnosticism of Huxley, but nor is it the theism of Paul Tillich or the atheism of Friedrich Nietzsche. Instead, given the consistent focus on hospitality to the stranger and the sacramental of the everyday, Kearney presents the potential for a new type of agnosticism alive not only to the possibility of finding God after rejecting God, but also finding redemption in the everyday. This anatheistic agnosticism is one forged in both the fires of atheism and the fires of theism: and it is all the stronger for it. Kearney does not claim anatheism as a form of agnosticism. He also denies that his work is that of a theologian, repeatedly distancing himself from the charge that anatheism is a third, Hegelian, position beyond atheism and theism. If anatheism is not a third position or a static position, it is an oscillation that involves elements of both atheism and theism in a tension that cannot be finally resolved. Ultimately, anatheism is a form of hospitality to both theism and

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Redeeming Agnosticism

211

atheism, discovering the sacred in the most unlikely places. As we have seen, this re-imagining of the sacred would be impossible without the resources of continental philosophy, just as it would be impossible without the theological resources of Christian, Jewish, and Islamic tradition. Kearney makes the case for discovering God within different atheisms and different religious traditions, revealing an anatheistic mode of pondering that cannot be limited to one concrete tradition or position. While anatheism seeks God in hospitality to the stranger and sacraments of everyday, anatheism is also a form of hospitality to the human experience of being agnostic. That it also might be an experience of redemption is not something about which we can be certain. But nor can that possibility be excluded without also rejecting the entirety of the anatheistic wager. Kearney is wagering everything on the idea that ‘in losing our faith, we may get it back again’.54 As I have argued elsewhere it is the lack of certainty which makes faith possible.55 While such language may sound utterly familiar and conventional to many theologians, the anatheist oscillation between atheism and theism ensures that the faith (that we may get back again) will always remain resolutely agnostic. That this agnostic faith is distinct from philosophical atheism, philosophical theism, or scriptural fundamentalism should not be in doubt. But that it might also be an iteration of the living faith of Christian theology is also surely possible.

BI B LI OGR APHY Anonymous. The Cloud of Unknowing, tr. Clifton Wolters (Middlesex: Penguin, 1961). Bonhoeffer, Dietrich. Letters and Papers from Prison: The Enlarged Edition (London: SCM, 1971). Collins, Guy. Faithful Doubt: The Wisdom of Uncertainty (Eugene: Cascade, 2014). Derrida, Jacques. ‘Circumfession’. Geoffrey Bennington and Jacques Derrida (Chicago: University of Chicago Press, 1993). Hart, Kevin. The Trespass of the Sign (New York: Fordham University Press, 2000).

54

Kearney (2011) p. 181.

55

Collins (2014).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

212

Agnosticism

Hooker, Richard. ‘Of the Laws of Ecclesiastical Polity’. In The Folger Library Edition of the Works of Richard Hooker, Vol. II W., edited by Speed Hill (Cambridge: Belknap Press, 1977). Kearney, Richard. The God Who May Be: A Hermeneutics of Religion (Bloomington, IN: Indiana University Press, 2001). Kearney, Richard. On Stories (London: Routledge, 2002). Kearney, Richard. Strangers, Gods and Monsters (London: Routledge, 2003). Kearney, Richard. Anatheism: Returning to God after God (New York: Columbia University Press, 2011). Kearney, Richard, and Jens Zimmerman (eds.). Reimagining the Sacred (New York: Columbia University Press, 2015). Le Poidevin, Robin. Agnosticism (Oxford: Oxford University Press, 2010). Moltmann, Jürgen. The Crucified God: The Cross of Christ as the Foundation and Criticism of Christian Theology (London: SCM, 1974). Pattison, George. Agnosis: Theology in the Void (London: Palgrave Macmillan, 1996). Ricoeur, Paul. Vivant jusqu’ à la mort (Paris: Seuil, 2007). Weatherhead, Leslie. The Christian Agnostic (London: Hodder and Stoughton, 1965).

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Index (Digital readers will find it useful to note that indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages.) Absolute 149–51 Absolute being 178–80 absolute knowledge 21, 143, 149–50, 152–3, 156–8, 160–1 Accidental being 178–9 Actual being 178–9 afterlife 108 agnostic belief 16–17, 54–5 absurdity of judging the conjunction 73 as an akratic state 66 justification 61–5 Moore’s paradox 69–70 rational or irrational 75–7 agnosticism apophatic agnosticism 22–4, 166 association with atheism 5–6, 186–7, 189 cancellation agnosticism 5 distinction from atheism 102–3 definitions 2, 4–5 epistemological agnosticism 88 existential agnosticism 33, 36 relationship with semantic agnosticism 35–6 fideistic agnosticism 166 first-order agnosticism 53–4 insouciant agnosticism 165 metaphysical context 3, 7–8 origin of term 7–8, 49 as a phase 190 philosophical agnosticism 16–19, 21, 163–4 positioned between theism and atheism 186–7 association with religious belief 20 association with theism 6–7, 13–14, 24, 85, 186–7, 189 distinction from theism 103–4 as third position 189–90 reconceived agnosticism 157–8 semantic agnosticism 14–17, 34–41 advantages of realism and fictionalism 44–5 strong agnosticism 5 akrasia 66

Albert the Great 8–9 Albertus Magnus 166–7, 183 ambivalence 92–3, 95n31 anatheism 25–6, 187–91, 194–5 ambiguity 193 approach of religions 192 Christian influence 202 eucharistic agnosis 203 influenced by Virginia Woolf 201 Jewish influence 196 redeeming agnosticism 208 sacramental aspect 207 sources 195 anathesitic wager 187, 192–5 Angel, Leonard 124–7 Anscombe, Elizabeth 167 Anselm 179–80 anti-realism 113 apophaticism 83–4, 163, 182, 188 Aquinas, Thomas 6–9, 22–4, 166–7, 170, 175, 178–8 apophatic dimension 182 Commentary on Boethius’s De Triniate 171 Commentary on the Sentences 171 influence on Anthony Kenny 167, 178–81 Summa Contra Gentiles 173 Summa Theologiae 174 Aquinas–Calvin (AC) model 87 Arendt, Hannah 196 Aristotle 121, 166–7, 171–2, 176–7, 196 atheism 52, 140–1 association with agnosticism 5–6, 186–7, 189 distinction from agnosticism 102–3 association with anatheism 192 debates on 1–2 Kenny’s criticism of 168–9 metaphysical problem 7–8 New Atheism 9–13 positive and negative 168 reading of Hegel 142–3 association with science 9–10

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

214 Avnur, Yuval 61n21 axiological transcendence (AT) 122–4, 126 being 178–9 see also existence belief approaches to agnosticism 82–3 episodic 99–101 beneficence of God 91 Bible 99n43 mentions of ‘mystery’ 93–4 Bonhoeffer, Dietrich 202 Bowden, H. S. 96–7 Braithwaite, Richard 37n4, 38–9 Brown, Frank Burch 96n31 Bullivant, Stephen 5–6 Bultmann, Rudolf 37n4 burden of proof 85 Butler, Judith 145 Byrne, Peter 114–15 Calvin, John 87 Caputo, John 202–3, 208 categorical belief 73, 76–7 Chesterton, G. K. 93n23, 98–9 Christian belief, agnosticism as a phase 24–5 Christian influence on anatheism 202 Clifford, W. K. 44–5, 82 Cloud of Unknowing, The (Anonymous) 199–201 Commentary on Boethius’s De Triniate (Aquinas) 171 Commentary on the Sentences (Aquinas) 171 Common being 178–80 conceptual possibility of agnostic belief 55–6 consciousness, totality of 160 continental philosophy 187–9, 195–6, 210–11 continuum hypothesis 59–60 Cupitt, Don 6–7, 37n4, 83 Damascene 171–2 Darwin, Charles 8, 96n31 Davies, Brian 170, 177–8 Dawkins, Richard 10–13 Demonstratio propter quid 176 Dennett, Daniel 10–11, 165 Derrida, Jacques 188–9, 197–200 deconstruction 188

Index Descartes, René 140 Desmond, Adrian 7–8 Diller, Jeanine 112–13, 117–19 Dionysian tradition 183 Dionysus 171–2 disjunctive semantics 40–1, 44 disjuncts 134–5 Divers, John 34n2 divine simplicity 168 divinity 58–9 Dominicanism 182–3 Dostoevsky, Fyodor 7n16, 195 doxastic attitudes 53 Draper, Paul 112n17 dualism 143–4 duality model 183–4 Dulles, A. 94n27 Eagleton, Terry 12–13 Eckhart, Meister 166–7, 183, 188–9 ego-release 126 Elliott, James 19–20, 118–19, 121–4, 127–30, 134–5 Ellmann, Richard 99n42 embodied moments 207 empty chalice 205 enlightenment 147–9, 158 entailment 17–19 episodic belief 99–101 epistemic justification 64 epistemic perspective 50 epistemological claim 163–4 Eshleman, Andrew 37n4 essence vs. existence 179–80 eucharistic agnosis 203 evidence 48–50 belief in insufficiency 51–3 of God’s existence 86 incompatible with faith 6 evolution, debate on 8, 11–12 evolutionary theory 169 ex poste justification vs. ex ante justification 67–8 existence 175 expressivism 37–9 faith 134–6, as form of life 83 incompatible with evidence 6 see also non-doxastic faith Fictional being 178–9 fictionalism 37–40

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Index advantages 44–5 knowledge of 43–4 vs. realism 39–43 truth-conditions 42–3 fideism 21, 48–9 fixity 146–7 fluidity of thought 146 form of life, faith as 83 formulae of sexuation 152–3 Forrest, Peter 109, 116–17 foundationalism 56, 61–2 ‘Four Horsemen’ 10–11 Frege, Gottlob 178 Friedman, Jane 48n1 friendship 101 Garvey, Brian 11–12 Geach, Peter 177–8 Gendler, Tamar Szabó 55n8 God conception of 12–13 definitions 83–4 knowledge of 173–7 as ‘qui est’ 171 rejection of 199 and the soul 197 and suffering 196–7 suffering of 202–3 transcendence 58 God’s existence 175 vs. essence 179–80 evidence 86 mystery of 93 preference for belief 91 sensus divinitatis 87 trust in 101 types of ‘being’ 179–81 unknowable 6–7 according to Wittgenstein 155–6 God’s nature, unknowable 6–7 Goldman, Alvin I. 67–8 Grayling, Anthony Clifford 9n23 Great Pumpkin example 85–6 Greenberg, Irving 196 Gutting, Gary 17–18, 88–9, 91–2, 95 happiness 176–7 Harris, Sam 10–11 Harrison, Peter 8–9 Hart, Kevin 188–9 Hawthorne, John, Daniel Rothschild, and Levi Spectre 72n31

215

Hegel, G. W. F. 21–2, 140–52, 156–7 comparison with Wittgenstein 157 Hick, John 112–13 Hillesum, Etty 196 Hinduism 124–5 Hitchens, Christopher 10–11 Holocaust 196–7 Hooker, Richard 204–7 hopeful ambivalence 17–19 Hovda, Paul 68–9 human cognition 172–3 Hume, David 56–7, 163 Huxley, Thomas 7–8, 11–12, 44–5, 49, 140–1, 208–10 idealism 113, 115–16 vs. realism 154 Identical being 178–9 ietsism 19–20, 137 vs. ultimism 117 imaginative fulfilment 133–4 indifference 92–3 individual agnosis 207 Individual existence 178–80 Ineffable 153–4 Insole, Christopher J. 113 Intentional being 178–9 International Social Survey Programme (2008) 2n4 James, William 44–5, 58–9, 62–4, 82, 89–91, 95, 99n43, 100–1 John of Damascus 171 John Paul II (Pope) 94n28 Judaism and anatheism 196 Jung, C. G. 177 justification, ex poste vs. ex ante 67–8 Kant, Immanuel 6–7, 19–20, 56–7, 143–5, 149, 163, 166 religious pluralism 112 Kearney, Richard 25–6, 186–95 background 191–2 eucharistic agnosis 203 individual agnosis 207 influenced by Dietrich Bonheoffer 202 influenced by Virginia Woolf 201 Jewish influence 196 redeeming agnosticism 208 sources of anatheism 195 Keller, Catherine 199–200, 202–3

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

216

Index

Kenny, Anthony 22–4, 167, 177–8 influenced by Aquinas 178–81 Kerr, Fergus 142–3, 151, 177–8 Kierkegaard, Søren 6, 20–1, 166 knower and known 143–6 knowing 140–2 self 146 and unknowing 22 knowledge, as norm of belief 68–9 Lacan, Jacques 152–3 Lane, Christopher 5 Langer, Susanne 99n43 Last Supper 204 Le Poidevin, Robin 37n4, 111n14, 164, 189n3 Levinas, Emmanuel 145–6, 196 lifestyle of an agnostic 18–19 Linus (Peanuts character) 85–6 logical positivism 168 logical possibility of agnostic belief 55 Lonergan, Bernard 167 Mackie, John 34–5 Mahaffy, J. P. 96–7 Marion, Jean-Luc 182–3 Martin, Michael 5–6, 11n37 materialism, vs. naturalism 8–9 Mawson, T. J. 7n16 McCabe, Herbert 12, 177–8 McDermott, Timothy 177 Melville, Herman 88 metaphysical context 3, 7–8 metaphysical naturalism 8–9 metaphysical transcendence (MT) 122–4, 126 methodological agnosticism 8–9 methodological atheism 8–9 methodological naturalism 8–9 miracles 58 modal discourse 14–15, 34 modal judgements 34 modernity 140 Moltmann, Jürgen 202–3 Monton, Bradley 48n1 Moore’s paradox 69–70 Moore-paradoxical sentences 70 moral discourse 14–15 moral judgements 34–6 Maimonides, Moses 175–6 multidimensional arguments 88–9 mundane experience 124–5

mystery 93 mystical atheism 198–200 mystical experience 131 vs. religious experience 130 mystical naturalism 124 mysticism 125–6, 183 Nagel, Thomas 22, 159 naturalism 11–12, 124, 136, 169 vs. materialism 8–9 negative atheism 168 as agnosticism 5–6 negative theology 188 neo-Darwinianism 169 new agnosticism 109, 142–3 New Agnosticism 14–17, 19 New Atheism 9–13 Nietzsche, Friedrich 210 non-All (pas-tout) 152–3 non-believers 2 see also atheism non-categorical belief 76–8 non-doxastic faith 19–20, 107–9, 112–14, 120–2, 134–7 non-ultimistic religious conceptions 120 Notion 149–50 not-knowing 186–7, 194–5 noumenon 143–4, 149 Nozick, Robert 133 objective self 159 Old Testament 99n43 mentions of ‘mystery’ 93–4 ontological pluralism 125–6 otherness 192, 197–9, 207 parallax gap 148 Pascal, Blaise 89–91, 95, 100–1, 182 Peanuts comic strip 85–6 permanent agnosticism 5 Permanent Agnosticism in Principle (PAP) 11 Phillips, D. Z. 37n4 phronesis 196, 199 physicalism 125–6 Plantinga, Alvin 7n16, 56n10, 85–9, 95, 103–4 Plato 166–7, 180, 205 pluralism 19–20, 108, 112, 125–6 pluralism hypothesis 112–15 pluriform vitality of interpretation 157–8

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

Index politics of practice 199–200 positive atheism 5–6, 11–12, 168 positivism 23, 37–9, 168–9 possibility of the impossible 210 Possible being 178–9 practical postulates 56–7 Predicative being 178–9 psychoanalytic thought 152–3 psychological dilemma 89–90 psychological possibility of agnostic belief 56–7 Pugh, Matthew 180–1 pure scepticism 110–11, 114, 132–4 ‘qui est’ 171 quidditative knowledge 176 Quine, Willard van Orman 169 Quintus, John Allen 98 Radical Orthodoxy 166 Real 114 realism 15–16, 36–8, 113, 115–17 advantages 44–5 vs. fictionalism 39–43 vs. idealism 154 reality 141, 143–4 Reformation 166 religion vs. science 8–10 religious belief associated with agnosticism 20 see also theism religious commitment 20 religious enquiry 111–12 religious experience 119 vs. mystical experience 130 as special experience 130–1 transcendence 132–3 ultimistic 131–2 religious faith, incompatible with evidence 6 religious pluralism 19–20, 108, 112 religious revelation 147–9, 158 representation of the world 144 restlessness of agnosticism 88 Revelation 166 Ribeiro, Brian 56nn9–10 Ricoeur, Paul 191–2, 203–5 Rocca, Gregory 172 rocket at the bottom of the garden 30, 45 Rose, Gillian 146–7, 150 Rosenkrantz, Sven 51n4 Rottschaefer, William A. 117n29

217

Rublev, Andrei 205 Russell, Bertrand 11–12, 51–2, 85, 99n43, 102–3 sacraments 204–7 sceptical agnosticism 5, 22–4, 165, 167 scepticism 147 relationship with agnosticism 164 Schellenberg, John 19–20, 59, 79–80, 107–9, 113–14, 116–25, 127–30, 132–4, 136 science vs. religion 8–10 scientific knowledge (Scientia; Aristotle) 176–7 Scientific Revolution 94–5 scientism 169 scope of agnosticism 164 second-order agnosticism 51 associated with agnostic belief 55–7 justification 57 secular hypotheses 59 as a form of truth agnosticism 41–3 relationship with truth and existential agnosticism 35–6 sensus divinitatis 87 Sextus 56–7 Shanks, Andrew 142–3 soteriological transcendence (ST) 122–4, 126–7 soul and God 197 special experience 130–1 Specific existence 178–80 Spinoza, Baruch 83 St Thomas see Aquinas, Thomas Fitzjames Stephen, James 89 strangers 192, 197–9 Stump, Eleonore 170, 183–4 Substantial being 178–9 suffering and God 196–7 of God 202–3 Summa Contra Gentiles (Aquinas) 173 Summa Theologiae (Aquinas) 174 Taoism 124–5 Taves, Ann 130–1 teapot argument (Russell) 11–12, 51–2, 85 temporary agnosticism 5

OUP CORRECTED PROOF – FINAL, 25/8/2020, SPi

218

Index

Temporary Agnosticism in Practice (TAP) 11 theism association with agnosticism 6–7, 13–14, 24, 85, 186–7, 189 distinction from agnosticism 103–4 conception of God 12–13 lack of certainty 7 as ultimism 109–10 see also religious belief theology 22–4 semantic agnosticism 36–41 Tillich, Paul 210 totality of consciousness 160 totalizing perspective 159–60 transcendence 107n2, 118–19, 122–4, 126–30, 136–7, 209 of a divinity 58–9 of God 58 religious experience 132–3 transmundane experience 124–5, 127–30 trust 101 Truth 149 truth agnosticism 33, 36 relationship with semantic agnosticism 35–6 semantic agnosticism as a form of 41–3 truth-conditions 34–5 fictional 42–3 fictive intent 43–4 theological 36–41 Ultimacy 126 Ultimate 122, 133–4 ultimism 19–20, 59, 107–9 vs. ietsism 117

vs. religious pluralism 112 ‘uninvited guest’ 192 unknowing 140–1, 143, 146–7, 149–50, 152–5 and knowing 22 Van Leeuwen, Neil 55n8 Vatican I 93–4 wager arguments 89 wagers 193–4 see also anatheistic wager wanderer metaphor 98–9 Ward, William 97 weak agnosticism 5 Weatherhead, Leslie 7n16 Western philosophy 143–4 Westphal, Merold 205–6 White, Roger 56n10 White, Victor 177–8 Whole of facts 153 Wiesel, Elie 197 Wilberforce, Samuel 8 Wilde, Oscar 96 Williams, Rowan 142–3, 156–8 Wittgenstein, Ludwig 6, 20–2, 37n4, 83, 140–4, 151–7, 167 comparison with Hegel 157 Wolf, Gary 93n23 Woolf, Virginia 201 Wordsworth, William 86 worship 127–8 Wright, Crispin 56n10 Zenk, Thomas 10 Zimmerman, Jens 205–7 Žižek, Slavoj 142–3, 148, 152–4, 158