The Continuum Companion to Hume 9781441114617, 1441114610

Citation preview

The Continuum Companion to Hume

9780826443595_Prelims_Final_txt_print.indd i

2/9/2012 9:38:07 PM

9780826443595_Prelims_Final_txt_print.indd ii

2/9/2012 9:38:07 PM

THE CONTINUUM COMPANION TO HUME

GENERAL EDITORS Alan Bailey Dan O’Brien

9780826443595_Prelims_Final_txt_print.indd iii

2/9/2012 9:38:07 PM

First published in 2012 by Continuum International Publishing Group The Tower Building 11 York Road London SE1 7NX

80 Maiden Lane Suite 704 New York, NY 10038

Companion to Hume EISBN 978 1 44111 461 7 © Continuum, 2012

British Library Cataloguing-in-Publication Data A CIP record of this title is available from the British Library

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any way or by any means, electronic, mechanical, photocopying, recording or otherwise, without the written permission of the copyright holder.

Typeset in Sabon by Newgen Publishing and Data Services Printed and bound in Great Britain

9780826443595_Prelims_Final_txt_print.indd iv

2/9/2012 9:38:07 PM

CONTENTS

List of Contributors Abbreviations for Works Written by Hume Acknowledgements David Hume – A Timeline INTRODUCTION

vii ix xi xiii 1

1. HUME’S LIFE, INTELLECTUAL CONTEXT AND RECEPTION

20

Emilio Mazza

2. HUME’S EMPIRICIST INNER EPISTEMOLOGY: A REASSESSMENT OF THE COPY PRINCIPLE

38

Tom Seppalainen and Angela Coventry

3. HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

57

Peter Millican

4. THE PSYCHOLOGY AND EPISTEMOLOGY OF HUME’S ACCOUNT OF PROBABLE REASONING

104

Lorne Falkenstein

5. CAUSATION AND NECESSARY CONNECTION

131

Helen Beebee

6. HUME ON SCEPTICISM AND THE MORAL SCIENCES

146

Alan Bailey

7. THE SELF AND PERSONAL IDENTITY

167

Harold Noonan

8. ‘ALL MY HOPES VANISH’: HUME ON THE MIND

181

Galen Strawson

9. ACTION, REASON AND THE PASSIONS

199

Constantine Sandis v

9780826443595_Prelims_Final_txt_print.indd v

2/9/2012 9:38:07 PM

CONTENTS

10. FREE WILL

214

James A. Harris

11. HUME ON MIRACLES

227

Duncan Pritchard and Alasdair Richmond

12. DAVID HUME AND THE ARGUMENT TO DESIGN

245

Andrew Pyle

13. PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF

265

David O’Connor

14. HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT

279

Julia Driver

15. HUME AND THE VIRTUES

288

Dan O’Brien

16. HUME’S HUMAN NATURE

303

Russell Hardin

17. HUME AND FEMINISM

319

Lívia Guimarães

18. HUME ON ECONOMIC WELL-BEING

332

Margaret Schabas

19. ‘OF THE STANDARD OF TASTE’: DECISIONS, RULES AND CRITICAL ARGUMENT

349

M. W. Rowe

20. HUME ON HISTORY

364

Timothy M. Costelloe

21. HUME’S LEGACY AND THE IDEA OF BRITISH EMPIRICISM

377

Paul Russell

Bibliography Index of Names Index of Topics

396 437 441

vi

9780826443595_Prelims_Final_txt_print.indd vi

2/9/2012 9:38:07 PM

LIST OF CONTRIBUTORS

Lorne Falkenstein Professor of Philosophy Department of Philosophy University of Western Ontario Canada

Alan Bailey Visiting Lecturer in Philosophy School of Law, Social Sciences and Communications University of Wolverhampton UK

Lívia Guimarães Lecturer in Philosophy Department of Philosophy Universidade Federal de Minas Gerais Brazil

Helen Beebee Professor of Philosophy Department of Philosophy University of Birmingham UK

Russell Hardin Professor of Politics Department of Politics New York University USA

Timothy Costelloe Associate Professor of Philosophy Department of Philosophy College of William and Mary USA

James Harris Senior Lecturer in Philosophy Department of Philosophy St Andrews University UK

Angela Coventry Associate Professor of Philosophy Department of Philosophy Portland State University USA

Emilio Mazza Associate Professor Institute of Human Sciences, Language and Environment Università IULM, Milan Italy

Julia Driver Professor of Philosophy Department of Philosophy Washington University in St Louis USA vii

9780826443595_Prelims_Final_txt_print.indd vii

2/9/2012 9:38:08 PM

LIST OF CONTRIBUTORS Peter Millican Gilbert Ryle Fellow and Professor of Philosophy Hertford College University of Oxford UK Harold Noonan Professor of Philosophy Department of Philosophy University of Nottingham UK Dan O’Brien Senior Lecturer in Philosophy Department of History, Philosophy and Religion Oxford Brookes University UK David O’Connor Professor of Philosophy Department of Philosophy Seton Hall University USA Duncan Pritchard Professor of Philosophy Department of Philosophy University of Edinburgh UK Andrew Pyle Reader in Philosophy Department of Philosophy University of Bristol UK Alasdair Richmond Senior Lecturer in Philosophy

Department of Philosophy University of Edinburgh UK M.W. Rowe Senior Lecturer in Philosophy Department of Philosophy University of East Anglia UK Paul Russell Professor of Philosophy Department of Philosophy University of British Columbia Canada Constantine Sandis Reader in Philosophy Department of History, Philosophy and Religion Oxford Brookes University UK Margaret Schabas Professor of Philosophy Department of Philosophy University of British Columbia Canada Tom Seppalainen Associate Professor of Philosophy Department of Philosophy Portland State University USA Galen Strawson Professor of Philosophy Department of Philosophy University of Reading UK

viii

9780826443595_Prelims_Final_txt_print.indd viii

2/9/2012 9:38:08 PM

ABBREVIATIONS FOR WORKS WRITTEN BY HUME

Abs An Abstract of a Book lately Published; Entituled, A Treatise of Human Nature Cited by paragraph and page number from the texts included in the editions of THN listed below. DIS A Dissertation on the Passions Cited by section and paragraph number from A Dissertation on the Passions; The Natural History of Religion, The Clarendon Edition of the Works of David Hume, ed. T.L. Beauchamp (Oxford: Clarendon Press, 2007). Hence DIS 1.2 = Section 1, paragraph 2. DNR

Dialogues concerning Natural Religion, ed. N. Kemp Smith, 2nd edn with supplement (London and Edinburgh: Thomas Nelson, 1947). Cited by part and page number (e.g. DNR 1.135 = Part 1, page 135). E

Essays, Moral, Political, and Literary, ed. E.F. Miller (Indianapolis: Liberty Classics, 1987). Cited by page number. EHU

An Enquiry concerning Human Understanding, ed. T.L. Beauchamp (Oxford: Oxford University Press, 1999). Cited by section and paragraph number, and supplemented by the corresponding page number in Enquiries concerning Human Understanding and concerning the Principles of Morals, ed. L.A. Selby-Bigge, 3rd edn with text revised and notes by P.H. Nidditch (Oxford: Clarendon Press, 1975). Hence EHU 5.1 / 40 = Section 5, paragraph 1 in the Beauchamp edition and page 40 in the Selby-Bigge edition. EPM

An Enquiry concerning the Principles of Morals, ed. T.L. Beauchamp (Oxford: Oxford University Press, 1998). Cited by section and paragraph number, and supplemented by the corresponding page number in Enquiries concerning Human Understanding and concerning the Principles of Morals, ed. L.A. Selby-Bigge, 3rd edn with text revised and notes by P.H. Nidditch (Oxford: Clarendon

ix

9780826443595_Prelims_Final_txt_print.indd ix

2/9/2012 9:38:08 PM

ABBREVIATIONS FOR WORKS WRITTEN BY HUME Press, 1975). Hence EPM 2.2 / 176–7 = Section 2, paragraph 2 in the Beauchamp edition and pages 176–7 in the Selby-Bigge edition. H

The History of England, from the Invasion of Julius Caesar to the Revolution in 1688, ed. W.B. Todd, 6 vols (Indianapolis: Liberty Classics, 1983). Cited by volume and page number (e.g. H1.155 = Volume 1, page 155). LDH Letters of David Hume, ed. J.Y.T. Greig, 2 vols (Oxford: Clarendon Press, 1932). Cited by volume, page number, and letter number (e.g. LDH 1.11–12, 5 = Volume 1, pages 11–12, letter 5). LFG

A Letter from a Gentleman to his Friend in Edinburgh, ed. E.C. Mossner and J.V. Price (Edinburgh: Edinburgh University Press, 1967). Cited by page number. MOL My Own Life. Usually cited by page number from Volume 1 of LDH above. NHR The Natural History of Religion. Cited by page or, where applicable, part and page number from Dialogues and Natural History of Religion, Oxford World’s Classics, ed. J.C.A. Gaskin (Oxford: Oxford University Press, 1993). Hence NHR 134 = page 134; NHR 1.135 = Part 1, page 135. NLH

New Letters of David Hume, eds. R. Klibansky and E.C. Mossner (Oxford: Clarendon Press, 1954). Cited by page and letter number (e.g. NLH 5–6, 3 = pages 5–6, letter 3). THN

A Treatise of Human Nature, eds. D.F. Norton and M.J. Norton (Oxford: Oxford University Press, 2000). Cited by book, part, section and paragraph number, and supplemented by the corresponding page number in A Treatise of Human Nature, ed. L.A. Selby-Bigge, 2nd edn with text revised and notes by P.H. Nidditch (Oxford: Clarendon Press, 1978). Hence THN 1.4.7.1 / 263 = Book 1, part 4, section 7, paragraph 1 in the Norton edition and page 263 in the Selby-Bigge edition. References to the Appendix of the Treatise make use of the abbreviation App. and are then given by paragraph and page number (eg. THN App. 1 / 623 = paragraph 1 of the Appendix in the Norton edition and page 623 in the Selby-Bigge edition). References to the editorial material of the Oxford Philosophical Texts or Clarendon Critical Edition versions of the Norton edition are by page number and the abbreviations THN-P and THN-C respectively.

x

9780826443595_Prelims_Final_txt_print.indd x

2/9/2012 9:38:08 PM

ACKNOWLEDGEMENTS

Special thanks to our copy editor, Merilyn Holme, for coaxing and prodding the book to completion; a far from easy task particularly given, in the final months, the looming festive season. We are immensely grateful to her and to Sarah Douglas at Continuum for commissioning the project. Thanks are due as usual to the Edwardian Tea Room in the Birmingham Museum and Art Gallery for refreshments and scholarly ambience, and Birmingham Central Library provided from its stacks a range of books and journals that would have done credit to a major university library. We would also like to take this opportunity to express our gratitude to the following colleagues at Oxford Brookes University, the University of Wolverhampton and Keele University: Stephen Boulter, Mark Cain, Beverley Clack, Geraldine Coggins, Meena Dhanda, Giuseppina D’Oro, Cécile Hatier, William Pawlett and Constantine Sandis. A. B. and D. O’B. Much of my work on this book has been timetabled around the ongoing DIY house project that Lucy and I are undertaking. This is something that I am sure Hume would appreciate. If not hands-on, he was certainly no slouch when it came to the upkeep of the home. When the ‘[p]laister broke down in the kitchen’ in his house in James’s Court, Edinburgh, he tells us that: [his repairman] having thus got into the house, went about teizing Lady Wallace [Hume’s tenant], and telling her, that this and the other thing was wrong, and ought to be mended. She told him, as she informs me, that everything was perfectly right, and she wou’d trouble the Landlord for nothing. Yet the Fellow had the Impudence to come to me, and tell me that he was sent by Lady Wallace to desire that some Stone Pavement under the Coal Bunker shoud be repair’d. I, having a perfect Confidence in Lady Wallace’s Discretion, directed him to repair it, as she desir’d. Having got this Authority, which cannot be good as it was obtain’d by a Lye, he not only pav’d the Bunker anew, but rais’d a great deal of the other Pavement of the Kitchen and laid it anew, nay, from his own head, took on him to white-wash the Kitchen: For all which, he brought me in an account of 30 Shillings. (NLH 115, 206) The contributors to this book may have explored Hume’s contributions to metaphysics, morality, religion and epistemology, but I have sympathetic appreciation of his knowledge of

xi

9780826443595_Prelims_Final_txt_print.indd xi

2/9/2012 9:38:08 PM

ACKNOWLEDGEMENTS that great human pursuit of house-building. And my deepest gratitude goes to Lucy my fellow plasterer, drywaller, spark and plumber and to Dylan who is still ‘patiently’ waiting for the kitchen to be finished. The skirting boards will be attached next week . . . D. O’B. Particular thanks on my behalf go to Ross Singleton for his longstanding friendship and our many lengthy conversations about religion, philosophy and international politics. Linda Dai has patiently coped with my tendency to introduce comments about Hume into a quite excessive number of contexts, and her backing and encouragement have played a crucial role in allowing me to complete my contribution to this volume. I also wish to thank my mother Dorothy Bailey for her support during the writing and editing process. A.B.

xii

9780826443595_Prelims_Final_txt_print.indd xii

2/9/2012 9:38:08 PM

DAVID HUME – A TIMELINE

1702

Death of William III and the accession of Queen Anne

1704

Battle of Blenheim; Isaac Newton, Opticks

1707

Union of England and Scotland

1711

Hume born in Edinburgh

1712

Thomas Newcomen’s steam engine

1713

George Berkeley, Three Dialogues between Hylas and Philonous; Anthony Collins, Discourse of Free-Thinking

1714

Death of Queen Anne and the accession of George I

Hume’s father dies

Bernard Mandeville, Fable of the Bees 1715

Jacobite rebellion in Scotland; death of Louis XIV; Louis XV becomes King of France at the age of five

1718

Inoculation for smallpox introduced in England

1720

Collapse of the South Sea Bubble; Edmond Halley becomes Astronomer Royal at Greenwich

1722

Robert Walpole becomes the equivalent of a Hume and his brother John enrol modern British Prime Minister as students at the University of Edinburgh

1725

Work starts on Grosvenor Square, London; Francis Hutcheson, The Original of Our Ideas of Beauty and Virtue

1726

Jonathan Swift, Gulliver’s Travels

Hume begins a legal education

xiii

9780826443595_Prelims_Final_txt_print.indd xiii

2/9/2012 9:38:08 PM

DAVID HUME – A TIMELINE

1727

Death of George I and the accession of George II; Isaac Newton dies Robert Greene, Principles of the Philosophy of the Expansive and Contractive Forces

1728

John Gay, Beggar’s Opera

1729

Hume abandons the idea of becoming a lawyer; his ‘new Scene of Thought’

1733

Alexander Pope, Essay on Man

1734

Voltaire, Lettres philosophiques (Lettres anglaises)

Hume travels to London; takes up employment in Bristol as a merchant’s clerk; resigns or is dismissed; relocates to France and works on the Treatise

1735

John Harrison’s chronometer; William Hogarth, Rake’s Progress

Hume moves from Rheims to La Flèche

1736

Joseph Butler, Analogy of Religion

1737

Returns from France to London

1739

Publication of Books One and Two of A Treatise of Human Nature

1740

Start of the War of Austrian Succession; Frederick II (Frederick the Great) becomes King of Prussia

Book Three of the Treatise

George Turnbull, Principles of Moral Philosophy 1741

Samuel Richardson, Pamela

Volume I of Essays, Moral and Political published in Edinburgh

1742

Walpole falls from power; Handel’s Messiah premieres in Dublin

1745

Second Jacobite rebellion; von Kleist discovers the ability of the Leyden jar to store electrical charge

Hume fails to secure the Chair of Ethics and Pneumatical Philosophy at Edinburgh; becomes tutor in England to the Marquess of Annandale Hume’s mother dies

1746

Jacobite army is decisively defeated at Culloden

Hume’s appointment as a tutor comes to an acrimonious end Hume becomes Secretary to Lieutenant-General James St Clair; accompanies St Clair on a military expedition attacking the coast of Brittany

xiv

9780826443595_Prelims_Final_txt_print.indd xiv

2/9/2012 9:38:08 PM

DAVID HUME – A TIMELINE

1748

War of Austrian Succession concludes; Treaty of Aix-la-Chapelle Excavations begin at Pompeii; Montesquieu, L’Esprit des lois; La Mettrie, L’Homme machine

Travels with St Clair on a diplomatic mission to Vienna and Turin Publication of Philosophical Essays concerning Human Understanding

1749

Buffon, first volumes of Histoire naturelle; David Hartley, Observations on Man; Henry Fielding, Tom Jones

Returns to Scotland and resides with his brother and sister at the family home in Ninewells

1751

Diderot and d’Alembert, Volume I of L’Encyclopédie

Hume moves to Edinburgh and is later joined by his sister Katherine

Adam Smith becomes Professor of Logic at the University of Glasgow

Publication of An Enquiry concerning the Principles of Morals

Adoption in Britain of the Gregorian calendar

Hume is unsuccessful in his candidacy for a chair at the University of Glasgow

1752

Elected Librarian to the Faculty of Advocates in Edinburgh 1753

British Museum founded

1754

Publication of Volume I of The History of Great Britain

1755

Lisbon earthquake; Samuel Johnson, Dictionary of the English Language

1756

Seven Years’ War

1757

Robert Clive and the East India Company are victorious at the Battle of Plassey

Hume resigns from his post as Librarian

Edmund Burke, A Philosophical Enquiry 1758

Philosophical Essays published under the new name of An Enquiry concerning Human Understanding

1759

Laurence Sterne, Tristram Shandy; William Wilberforce is born

1761

The Bridgewater Canal opens from Worsley Madame de Boufflers’s initial letter to Manchester to Hume

1762

Catherine II becomes Empress of Russia; Sarah Scott’s novel of a female utopian community, A Description of Millenium Hall

1763

Seven Years’ War concludes; Peace of Paris Catherine Macaulay, first volume of her History of England

Hume accompanies Lord Hertford to Paris and takes on the duties of Secretary to the British Embassy Hume and Madame de Boufflers meet for the first time

xv

9780826443595_Prelims_Final_txt_print.indd xv

2/9/2012 9:38:09 PM

DAVID HUME – A TIMELINE

1764

Thomas Reid, Inquiry into the Human Mind; Horace Walpole, The Castle of Otranto Mozart’s Symphony No. 1 composed in London during the family’s European tour

1765

Matthew Boulton finishes building the Soho Hume is officially confirmed as Manufactory in Birmingham Secretary shortly before his post comes to an end

1766

Immanuel Kant, Dreams of a Spirit-Seer Elucidated by Dreams of Metaphysics

Hume returns to London, accompanied by Jean-Jacques Rousseau; Rousseau accuses Hume of being part of a conspiracy against him A Concise and Genuine Account of the Dispute between Mr. Hume and Mr. Rousseau Hume spends the final months of the year in Ninewells and Edinburgh

1767

James Craig’s plan for New Town, Edinburgh is adopted; Royal Crescent, Bath, started

Travels to London to take up the office of Under-Secretary of State in the Northern Department

1768

James Cook’s first voyage to the Pacific

Hume retires from public office

1769

Josiah Wedgwood and Thomas Bentley open Hume returns to Edinburgh their Etruria factory near Stoke-on-Trent James Watt’s steam engine, James Hargreaves’ spinning jenny, and Richard Arkwright’s water frame are patented Arthur Wellesley, later Duke of Wellington, is born in Dublin

1770

The future Louis XVI marries Marie Hume has a house built for himself Antoinette; Baron d’Holbach, Système de la in Edinburgh’s New Town; rumours nature reach Paris that Hume might be about to marry Nancy Orde

1771

Tobias Smollet, Humphry Clinker

1773

Boston Tea Party

Hume and his sister move into the new house – St Andrew Square, off St David Street

Hester Chapone, Letters on the Improvement of the Mind; Oliver Goldsmith, She Stoops to Conquer

xvi

9780826443595_Prelims_Final_txt_print.indd xvi

2/9/2012 9:38:09 PM

DAVID HUME – A TIMELINE

1774

Death of Louis XV; Louis XVI becomes King of France Joseph Priestley discovers ‘dephlogisticated air’ (oxygen)

1775

John Wilkinson’s cannon-boring machine; Jane Austen born

The ‘Advertisement’ repudiating the Treatise

1776

Declaration of American Independence

Hume dies in Edinburgh

Edward Gibbon, Decline and Fall of the Roman Empire; Adam Smith, The Wealth of Nations; Thomas Paine, Common Sense; Jeremy Bentham, Fragment on Government 1777

Publication of Life of David Hume, written by Himself

1779

World’s first iron bridge is completed across Publication of the Dialogues the River Severn at Coalbrookdale concerning Natural Religion

1782

Atheism openly avowed in print in Britain for the first time – Matthew Turner, Answer to Dr Priestley’s Letters to a Philosophical Unbeliever

1783

Peace of Versailles establishes independence of American colonies

1785

Edmund Cartwright patents his power loom

1787

Founding in Britain of the Committee for the Abolition of the African Slave Trade

1789

French Revolution

xvii

9780826443595_Prelims_Final_txt_print.indd xvii

2/9/2012 9:38:09 PM

9780826443595_Prelims_Final_txt_print.indd xviii

2/9/2012 9:38:09 PM

INTRODUCTION

metaphysics and epistemology. There is, in fact, an overwhelming case for saying that no other eighteenth-century writer’s account of English history came close to matching the intellectual quality and non-partisan nature of Hume’s own narrative, and in this particular case those genuine merits were, for once, rewarded by the approbation of substantial sections of the public. Hume’s current reputation is, therefore, something that stands in need of explanation. How has an eighteenth-century Scottish intellectual and writer who enjoyed his greatest success amongst his contemporaries as a historian, economist and writer of polite essays arrived at the status and, in the eyes of the editors of this volume, the wholly deserved status of being viewed as the greatest British philosopher? In many respects the answer lies in the very features of his philosophical writings that saw them subjected to so much criticism when Hume was alive. Epistemological scepticism, even of a radical variety, is no longer seen as constituting any kind of threat to morality and social order; so it is now possible to respond to the sceptical arguments deployed within Hume’s writings as providing us with a series of fascinating puzzles that may succeed in pointing us towards important

In 2005 the British Broadcasting Corporation ran a poll asking Radio Four listeners to say whom they regarded as the greatest philosopher of all time. Such familiar philosophical luminaries as Plato, Aristotle, Descartes and Kant all featured prominently in the subsequent voting, and Marx’s immense influence within the political arena saw him, rather predictably, taking first place. However, the pre-eminent British philosopher, and the philosopher with the second highest overall number of votes, was David Hume. In his own lifetime Hume certainly possessed a substantial reputation as a public intellectual. In some respects, though, it would be more appropriate to talk in terms of his notoriety rather than his reputation. His supposedly sceptical epistemological views and the manner in which his writings seemed to develop a series of pointed criticisms of religious belief attracted vituperative criticism from many of his contemporaries. It is also a striking fact that much of his fame sprang from his ostensibly non-philosophical writings. Until his death in 1776 Hume enjoyed a great deal of influence as a writer on matters of economics. Moreover, sales of his History of England made him independently wealthy and brought him to the attention of far more readers than were interested in works of 1

9780826443595_Intro_Final_txt_print.indd 1

2/9/2012 9:37:57 PM

INTRODUCTION these staples of theistic apologetics. Just as significantly, however, Hume’s evident willingness to question religious dogma at a time when the social and cultural pressure towards internalizing such beliefs was so strong marks him out as a person who was prepared to be guided by argument and the available evidence instead of suppressing his critical faculties at the behest of superstition and the power-structures of religious authority. In this respect, Hume amply meets the essential requirement that a true philosopher, a philosopher of genuine integrity, must answer only to the autonomous demands of the reflective intellect. It also seems to be the case that once Hume’s epistemological and irreligious views are no longer predominantly seen as views that need to be repudiated as aggressively as possible, other valuable aspects of his philosophical outlook become increasingly easy to recognize. Given the disappointing results of attempts at a priori metaphysics, Hume’s denunciations of the application of a priori reasoning outside the sphere of issues of ‘quantity and number’ (EHU 12.27 / 163) seem amply vindicated by the historical record. Thus philosophical inquiry needs an alternative methodology if it is not simply to repeat past errors in ever more complex forms. And Hume’s ‘experimental’ method, with its commitment to being guided by experience, seems to meet this need. There might perhaps be some worries that this approach actually amounts to a simple repudiation of philosophy in favour of the investigative methods of the sciences. In Hume’s hands, however, it constitutes not an abandonment of philosophy but a confirmation that at least some philosophical conundrums can be satisfactorily dissolved by paying due attention to the empirical facts. Confronted, for example, by the question

truths about the nature of philosophy or the incoherence of certain aspects of our selfconception as inquirers and agents without those arguments constituting instruments of intellectual self-annihilation. Moreover, once this fear of sceptical conclusions has been dissipated, it becomes psychologically easier to acknowledge the inadequacy of so many of the standard supposed refutations of sceptical arguments. Hume’s own recognition of the power of these arguments accordingly comes to be seen as compelling evidence of his own intellectual integrity and powers of analysis. This issue of intellectual integrity also has a bearing on present-day reactions to Hume’s criticisms of religion. Britain in the eighteenth century was an overwhelmingly Christian country, where overt expressions of disbelief could still attract substantial prison sentences and books regarded as attacking Christianity were frequently subjected to determined campaigns of suppression. Today, in contrast, there is substantial evidence that between 30 and 40 per cent of the British population do not believe in God or any Higher Power analogous to a person. And although the United States signally lags behind almost all Western European states in this regard, agnosticism and atheism are making some inroads even in that hitherto hostile environment. There is, accordingly, a far more receptive audience in the current climate for arguments challenging the metaphysical underpinnings of a religious world-view and the complacent supposition that religious convictions constitute crucial support for moral behaviour and an appreciation of the value of life. Hume’s writings provide such arguments in abundance, and his critiques of the argument to design and the credibility of testimony to alleged miracles still constitute some of the most trenchant attacks ever launched on 2

9780826443595_Intro_Final_txt_print.indd 2

2/9/2012 9:37:57 PM

INTRODUCTION whist, to avoid challenging him to any card games involving large sums of money. In a sense, of course, Hume’s personal virtues do not add to the importance of his intellectual achievements. But they do confirm one important thing, namely that the philosophical outlook embraced by Hume is one that is entirely compatible with a flourishing human life that combines generous concern for the well-being of other people with ample enjoyment of a full range of social and intellectual satisfactions. This combination of the power of Hume’s thought and the engaging quality of his personality has undoubtedly helped in bringing together the contributors to this volume. As editors, we were repeatedly pleasantly surprised by the enthusiasm expressed for this project by potential contributors, and we hope that the finished anthology succeeds both in illuminating Hume’s own achievements and in suitably showcasing the commitment to Humean scholarship manifested by all the authors represented in the following pages. Emilio Mazza opens the volume by drawing us into Hume’s world, one far from the ivory tower – a world of business, military expeditions, international diplomacy and Parisian ladies. But Le Bon David always sought refuge from this heady world in work, in friendships and in his pursuit of literary fame, his ‘ruling passion’. His academic legacy and fame, however, are perhaps rather fortunate given that he would have been happy to stay at home in the borders of Scotland, if his brother had not married, or to join the army if he had discovered its pleasures and camaraderie at a younger age. Mazza’s evocative biography illuminates a life of travel and friendships with a portrait of a cheery, avuncular man toddling around Edinburgh, being dragged out of a

‘What obligation do we have to obey this or indeed any government?’, we might initially be at a loss to know how to proceed. Science, we have been repeatedly told, cannot answer normative questions. Conceptual analysis, it is tempting to suppose, can at best clarify the sense of the question, and a priori reasoning of a non-mathematical kind cannot be relied upon to yield anything more substantial than vacuous tautologies. Hume’s account of human nature, in contrast, allows us to see this question as an idle one. There may indeed be scope for choosing which government to follow. But our psychological properties mean that some institutions of government will inevitably arise in all circumstances that are ever likely to persist for a significant length of time. Moreover, once these institutions have arisen, their success in securing high levels of obedience is equally inevitable irrespective of our normative speculations. It would be remiss, however, of any account of Hume’s well-merited appeal to present-day philosophers and anyone interested in understanding the place of human beings in the world to ignore the question of Hume’s personal character. Although this has frequently been traduced by defenders of religion and people who mistakenly suppose that seriousness of purpose must be evidenced by tortuous writing, pompous pretentiousness and a complete absence of humour, it is clear from the record of Hume’s life that he was a benevolent man of amiable temperament, a good and loyal friend, and a master of comic self-deprecation and subtle word-play. If one were planning a fantasy dinner party, it is difficult to imagine any philosopher in history who would make a more winning and entertaining guest or a more congenial host, though it would probably be advisable, given his reported skill at 3

9780826443595_Intro_Final_txt_print.indd 3

2/9/2012 9:37:57 PM

INTRODUCTION is the source of the vivacity of our experience, of its intentional content, and of the believability of our ideas. This account is contrasted with Descartes’s theory of ideas and with interpreters of Hume who see him as a proto-logical positivist. Peter Millican turns to Hume’s account of inductive reasoning and his ‘famous argument’ to the conclusion that we have no warrant for our beliefs about unobserved matters of fact. Millican spells out the steps of Hume’s argument as articulated in the Treatise, the Abstract and the first Enquiry. All beliefs concerning matters of fact are grounded in causal reasoning but, Hume argues, knowledge of causal relations cannot be acquired a priori, nor can it be gained via inductive reasoning. In place of such support Hume provides an account of belief grounded in custom or habit. However convincing Hume’s arguments may be, there is undoubted tension between his seemingly sceptical conclusions and his embrace of inductive science, his ‘experimental’ approach to the study of human nature, and his empirical approach to history. Some interpreters of Hume take him just to be concerned with a psychological description of thinkers and not with issues concerning justification and warrant. Millican, however, argues that Hume is interested in normative questions; it is important to be clear, though, on the target of Hume’s scepticism – and that is Locke’s conception of reason, what Millican calls his perceptual model. Such scepticism, however, does not engender what has come to be called ‘The problem of induction’. The purpose of Hume’s form of ‘mitigated’ scepticism is not, as with Descartes, to prompt us to discover a sure path to certain knowledge, but rather to instil in us a suitable level of modesty and caution concerning our epistemic practices. Furthermore, Hume’s naturalistic account

bog, saving a man from execution after that man’s unsuccessful attempt to commit suicide in Paris, and having deep and sometimes stormy relationships with the literati of his day including, amongst many others, Adam Smith, Rousseau, d’Alembert and Lawrence Sterne. We come away with the impression that Hume had a good life – one with much friendship, fame and fortune – and if one can ever say this, Hume also had a good death. To the end he was in good spirits, reading his beloved classics, and revising his Dialogues concerning Natural Religion. Tom Seppalainen and Angela Coventry take a ‘fresh look’ at Hume’s theory of ideas and impressions. The notion of liveliness or vivacity that distinguishes mere ideas from beliefs – beliefs being vivid ideas – is usually taken to be a phenomenological one and various interpretations of the nature of vivacity are considered. It has been characterized in terms of qualitative feel although Seppalainen and Coventry argue that thinking of perceptions in this way ignores their intentional content and the way perceptions seem to be of the world. An improvement, then, is to read the phenomenology of perception not in terms akin to those describing the intensity of colour in a picture, but in terms of ‘presentedness’ (or, according to another intentional reading of Hume, in terms of verisimilitude and the feeling of reality). Experience presents the world to one. Seppalainen and Coventry applaud such intentional readings, but they argue that Hume does not use vivacity to refer to the phenomenological qualities of individual perceptions, but rather to sequences of ideas and impressions. Only patterns of change can be vivid in the requisite sense. Our very notion of the existence of the external world depends on the constant and coherent flux of our perceptions and, they argue, such flux 4

9780826443595_Intro_Final_txt_print.indd 4

2/9/2012 9:37:57 PM

INTRODUCTION the usually apparent regularity of nature. And Hume’s account of the vivacity of belief can show how the strength of our beliefs depends on the uniformity of our experience. This account does not depend on a mathematical calculation of probabilities, but rather on associationist psychological processes of vivacity transfer. Falkenstein concludes by showing how Hume takes his scepticism as supporting his ‘system’ and his account of empirical reasoning. Helen Beebee heads into stormy water – into Hume’s account of causation, a hotly debated topic: the so-called New Hume Debate focusing on the question of whether Hume takes there to be real causal powers in nature – oomphs pushing billiard balls around tables – or whether he thinks that there are really only constant conjunctions and regular brute patterns in the world. Beebee investigates this question, concentrating on three claims to which Hume seems to be committed. First, Hume suggests that we project causal connections onto the world, connections that are not really there. Second, the concept of causation includes the notion of necessity. We think of causal connections as those that are necessarily connected together: the red ball must move off in that direction given that it was struck in that way by the white ball. And third, causal talk is objective in that we can talk correctly of causes and we can sometimes make false causal claims about the world. These commitments seem to be inconsistent since it is not obvious how causal talk can be objective when Hume does not think that there are really causal connections in nature and that they are projections of our cognitive processes. Beebee discusses various ways to resolve this (perhaps only apparent) inconsistency. The traditional interpretation of Hume claims that causation is just constant conjunction. The sceptical

provides an explanation of how human beings can – and actually do – reason inductively and this, given the impossibility of any rational foundations for such reasoning, provides us with all the support we require for our causal inferences. Lorne Falkenstein further explores Hume’s account of causal reasoning and the ‘system’ of the Treatise. It is constant conjunctions in experience that impel the imagination to form beliefs concerning causes and effects. Habits of thought guide our reasoning and not rational argument or judgement. As discussed by Coventry and Seppalainen, beliefs are seen as vivacious ideas, and Falkenstein stresses that Hume does not think of vivacity in terms of the intensity of an image. Beliefs, rather, amount to dispositions of the mind (for example, to incline one to act in certain ways and to focus one’s attention). Beliefs, then, are the product of the principle of association of causation, although Falkenstein also suggests how Hume might have included the relations of contiguity and resemblance in his account of reasoning. If, however, belief formation is just a matter of habit, how can it be so that some ways of forming beliefs are seen as better than others? Hume suggests that we follow certain general rules, those that are learnt from past experience, such as, like objects in like circumstances will have like effects. Particular inferences we make can be assessed against such general rules. This is Hume’s logic of causal inference. Such rules can also be extended to cover probable reasoning. It is not the case that we have uniform experience, but this does not lead to us rejecting such general rules. Instead, in cases where like causes do not seem to lead to like effects we come to believe – again via habit grounded in past experience – that there is some hidden cause that is disrupting 5

9780826443595_Intro_Final_txt_print.indd 5

2/9/2012 9:37:57 PM

INTRODUCTION radical position than mere fallibilism. In Bailey’s judgement, this scepticism is better interpreted as a stance that does not endorse any beliefs as possessing a positive degree of epistemic justification except for beliefs about very simple necessary truths that can be grasped without going through any process of inference and beliefs about the content of our present ideas and impressions. It is clear, however, that if such scepticism is an integral component of Hume’s thought, then it co-exists with Hume’s assent to a detailed and carefully constructed account of human nature that is supposed to be both true and useful. Even a moment’s reflection on the Treatise’s subtitle, which is Being an Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects, indicates that it would be a disastrous misreading of Hume’s views to construe him as simply a destructive sceptic. Bailey’s solution to this interpretative conundrum is to claim that Hume views radically epistemological scepticism and properly conducted empirical inquiries as mutually supportive. Hume thinks that sceptical arguments are indeed successful in placing us in a position where only our acceptance of the view that scarcely any of our beliefs are rationally justified can allow us to deny, without being guilty of bad faith, that such sceptical arguments provide rationally compelling grounds for that assessment of our beliefs. However, Bailey argues that Hume does not see this as posing any threat to the ability of our belief-forming mechanisms to generate and sustain in existence all the beliefs we need to guide our actions. Nor indeed does Hume view it as undermining our capacity to assent to relatively sophisticated scientific theories. Where such theories are constructed using systematized and reflective versions of common sense methods

realist, in contrast, sees Hume as accepting that there are causal powers in nature; it is just that we cannot come to have knowledge of them. The projectivist interpretation adopts a non-cognitivist stance: our claims concerning causal relations are subject to norms, but these norms are constituted not by features of the world independent of our judgements concerning its causal structure – by real causal powers in nature – but by certain ‘rules’ which we have come to appreciate concerning how we judge of causes and effects, rules that enable us to override errant judgements in particular cases. A clue to the correct interpretation can be found in Beebee’s claim that Hume is driven by his opposition to the Cartesian Image of God Hypothesis. There are two aspects to this hypothesis: we are, as the Bible says, made in God’s image, and we have epistemic abilities in line with such an origin. The nature of reality is accessible to human reason – we can come, through a priori reasoning, to have knowledge of the world and specifically of its causal structure. Beebee notes certain aspects of this picture in the sceptical realist approach and thus argues that this cannot be the correct interpretation of Hume. Of the remaining options, Beebee favours projectivism over the traditional interpretation. Alan Bailey then undertakes an examination of the equally vexed issue of the prospects for providing a unified account of Hume’s philosophical outlook that satisfactorily accommodates both his ambitions to construct a science of human nature and the sceptical elements of his thought. If Hume’s putative scepticism actually amounted to nothing more than a modest epistemological fallibilism, as some recent commentators have supposed, there would be no real tension to overcome here. Bailey argues, however, that Hume’s scepticism is a much more 6

9780826443595_Intro_Final_txt_print.indd 6

2/9/2012 9:37:58 PM

INTRODUCTION that is, that the soul or mind is a substance, be it physical or non-physical. All we can do is provide an account of what causes us to have the mistaken belief that there are enduring selves. Such an account includes certain identity-ascribing mechanisms of the imagination – those grounded in the principles of association of resemblance and causation – mechanisms that generate belief in the self as well as in the continued existence of the external world and of bodies. Hume, however, is dissatisfied with his conclusion. He thinks that he is committed to two inconsistent principles: that distinct perceptions are distinct existences, and that the mind never perceives any connection between them. Perplexingly, however, these principles are not inconsistent, and uncovering why Hume claims them to be so is a key difficulty for interpreters of Hume’s views on the soul and the self. Noonan suggests that Hume realizes that his account does not explain our continuing belief in personal identity. One can accept Hume’s empiricist conclusions with respect to the external world and give up the notion that there is a substance or substrata underlying the properties of bodies, but we cannot accept this with respect to the self. Why not? Hume did not know. Galen Strawson has a distinct account of why Hume’s ‘hopes vanish’. Hume discovers – late in the day, in the Appendix to the Treatise – that his whole empiricist philosophy depends on a conception of the mind that his empiricism does not allow him to have. His genetic account of our belief in an enduring self relies on the principles of association – it relies on the assumption that we have a ‘Principle-Governed Mind’. This explains our belief in the self as well as our belief in the external world and in causation.

of inquiry and are accordingly supported by experience and experiment, sceptical discoveries are incapable of preventing us from giving our assent. And where those theories are not supported by experience and experiment, Hume can, as an empiricist, rejoice in their destruction. Thus Bailey holds that Hume sees scepticism and the proper experimental method of inquiry working in tandem. Sceptical arguments curb the power of the imagination to generate beliefs that are not the products of the observed correlations that give rise to causal inferences. And the experience-based beliefs towards which we accordingly gravitate generate a plausible picture of the workings of the human mind that makes it even more difficult for us to represent ourselves as capable of arriving at many beliefs that genuinely qualify as rationally justified. Harold Noonan and Galen Strawson both explore what Hume calls the labyrinth of personal identity. Noonan considers various arguments in Hume against the Cartesian conception of personal identity, against, that is, the existence of an enduring self, identical from moment to moment and from day to day. Hume’s empiricism demands that we have an impression of such an entity, but this we do not have – all we find, on introspection, is a bundle of variously related perceptions. Hume’s ‘master-argument’ establishes that all perceptions are logically capable of an independent existence. There is thus no need for the ‘unintelligible chimera of substance’ (THN 1.4.3.7 / 222) in which properties must inhere. This is so for physical things, such as wax – contra Descartes, wax for Hume is just a collection of properties – and for human beings: we do not require an enduring soul underlying our ever-changing properties. Hume thus criticizes a supposition of both materialists and immaterialists, 7

9780826443595_Intro_Final_txt_print.indd 7

2/9/2012 9:37:58 PM

INTRODUCTION Sandis also turns to interpretations of Hume’s famous claim that ‘[r]eason is, and ought only to be the slave of the passions’ (THN 2.2.3.4 / 414) and argues that the received Humean theory of motivation is unfounded. This is the view that an agent cannot be motivated by belief alone, but only by a belief along with an appropriately related desire. Sandis claims, though, that such an account is not to be found in Hume. It is also suggested that Hume does not equate belief and opinions with judgement. Ideas and beliefs are distinguished by their vivacity, and the vivacity of judgements should be seen as lying somewhere between that of ideas and beliefs. This is relevant to Hume’s account of morality: Hume does not talk of moral judgements but, on Sandis’s account, this still allows Hume to have an account of moral beliefs and of their motivational force. James Harris turns to liberty and necessity, and to what Hume calls ‘the most contentious question of metaphysics, the most contentious science’ (EHU 8.23 / 95). Hume is often thought to be an advocate of an early version of what is now called compatibilism, and it has been claimed that there is nothing distinctive about his position. Locke and Hobbes had also discussed this question and suggested compatibilist answers. Harris, however, argues that Hume is not just rehashing their arguments. Importantly, it is claimed, Hume is not a determinist in the modern sense, unlike, for example, Hobbes. Determinism is a metaphysical stance and Hume eschews metaphysical questions. His claim is not that we have reason to think that the laws of nature cannot change – that they are determined; everyday experience, rather, leads us to expect that people behave in regular ways and we interact with them in light of these regularities.

But such a conception of the mind goes beyond a loose association of distinct perceptions. In order to legitimately ground one’s philosophy in such an account of the mind it is required either that there is an observable real connection between the perceptions that make up the mind or grounds for claiming that such perceptions inhere in some kind of soul-substance. But Hume has argued against both possibilities. A possible response here is to take Hume as sheltering in his scepticism: the essence of the mind is unknowable to us and thus it cannot be this – the lack of knowledge of the Principle-Governed Mind – that leads him to despair. But, Strawson argues, such agnosticism cannot do the trick. Hume does need to, and does, assume a certain notion of the mind – a rule-governed one. He can perhaps remain agnostic about just how it works, but he cannot be agnostic about its very existence – and its very existence is what is incompatible with Hume’s empiricism. Strawson claims that Hume’s despair is a result of his acknowledgment of this deep inconsistency in his philosophy. Constantine Sandis explores Hume’s account of action and in so doing considers how reason, the will and the passions are related. Hume’s account of action is an empirical one: we acquire knowledge of a person’s reasons for acting from careful observation of human behaviour. This ‘science of man’ grounds Hume’s History of England and the study of this work highlights how Hume sees character as playing a key motivating role in our behaviour. Further, the historian is best placed to uncover the truth about human nature since he does not aspire to the detached perspective of the philosopher, nor is he too close to his subjects and thus prone to bias or the distorting influence of his particular interests and circumstances. 8

9780826443595_Intro_Final_txt_print.indd 8

2/9/2012 9:37:58 PM

INTRODUCTION the extent to which he and other irreligious thinkers of his time were forced to engage in misdirection and linguistic contortions in order to avoid social ostracism and the official suppression of their writings. The authors of the three chapters in this anthology that focus primarily on Hume and religion are therefore unanimous in presenting him as a rigorous and intellectually honest thinker who deploys a formidable set of arguments against any form of religious outlook based on the truth of theism or even a robust form of deism. Duncan Pritchard and Alasdair Richmond investigate Hume’s notorious arguments on the topic of the credibility of testimony concerning miracles. They are careful to locate these arguments within the broader framework of Hume’s reservations about our ability to justify expectations about the future in a rational, non-circular manner and his pragmatic response to those sceptical worries. Although causal reasoning cannot be supplied with a non-circular argumentative defence, it remains the case that human beings find such reasoning persuasive and continue to use it, even after exposure to sceptical arguments, as a touchstone for assessing whether particular beliefs are ones they are content to endorse or ones that are no more than mere foolishness. Consequently Pritchard and Richmond construe Hume as attempting to show that no testimony about the occurrence of miraculous events capable of serving as the foundations of a system of religion has ever met the standards of doxastic acceptability that normally prevail in less contentious cases when we are weighing human testimony concerning an alleged event against the implications of our observations of past natural regularities. Pritchard and Richmond reject the supposition that any form of a priori conceptual

Harris also notes various changes in emphasis between the Treatise discussion of this topic and that of the Enquiry. In the latter, Hume more squarely targets metaphysicians. Once we clarify the nature of liberty and necessity, long-running metaphysical and, in particular, religious disputes concerning God’s prescience and responsibility for evil will be undermined. Philosophy should ‘return, with suitable modesty, to her true and proper province, the examination of common life’ (EHU 8.36 / 103). Hume’s discussion of liberty and necessity is not a case of Hume engaging in metaphysical debate, the question then arising of whether his contribution is original or not – he is, rather, agnostic about all such issues, his discussion reflecting his empiricist attitude to questions concerning the regularity of human behaviour and morality. At this point the contributors turn to the subject of Hume’s views on the truth and utility of religious beliefs. In his own era he was interpreted as attacking Christianity and all forms of theistic belief. However, his arguments were frequently dismissed as inconsequential sophistries motivated not by a concern for the truth but by a desire to secure personal notoriety and increase the sales of his books. Such an assessment of the force of his arguments and his motivation for advancing them is now wholly discredited. Yet the recognition that he wrote on this particular topic in good faith and with a commitment to seeking the truth and promoting human well-being has led some present-day commentators to suggest that he was actually a defender of some philosophically purified form of theism that might even be compatible with the truth of Christianity. Such an interpretation seems to be based on nothing more substantial than his invariable courtesy when debating matters of religion and an almost inexcusable failure to recognize 9

9780826443595_Intro_Final_txt_print.indd 9

2/9/2012 9:37:58 PM

INTRODUCTION as confirmation of the truth of religious doctrines or teachings. If Hume is right to maintain that reports of alleged miracles fail to offer any genuine support for the bold claims advanced by religions about the ultimate nature of reality, where might such support be found? Hume’s religious contemporaries placed great confidence in the probative value of the design argument, and Hume undertook a detailed examination of this argument in his Dialogues concerning Natural Religion, which were published posthumously in 1779. Andrew Pyle accordingly presents in his chapter an overview of the complex discussion that occurs between Hume’s principal characters in the Dialogues, and he arrives at the conclusion that this work was intended to show that the design argument cannot legitimately support theistic conclusions and that a naturalistic explanation of the orderly nature of the universe is, when judged by everyday standards of good causal reasoning, more acceptable than a theistic one. An interpretation of the Dialogues along these lines might initially be thought to overlook Hume’s repeated suggestions that our intellectual faculties are wholly inadequate when confronted by the task of arriving at satisfactory conclusions about such rarefied matters of inquiry. Those pronouncements appear to give very strong support to the conclusion that Hume holds that the only legitimate response to questions about the ultimate origins of the universe is a stance of complete neutrality and suspension of judgement. However, Pyle draws an important distinction between a commitment to a particular hypothesis as more probable than all competing hypotheses with equivalently detailed content and a comparative judgement that a particular hypothesis is more likely to be

argument forms part of Hume’s case against belief in miracles, and they also maintain that it is a mistake to construe him as arguing that the kind of regularity in experience that would need to be interrupted in order for an event to qualify as a plausible candidate for being a miracle would be so well entrenched and confirmed that no possible amount of human testimony could render it appropriate to believe that an interruption had occurred. They emphasize that for Hume it is always a contingent matter whether the testimony offered is weighty enough to overcome the initial presumption that a hitherto well-confirmed natural regularity with no previously known exceptions has not abruptly come to an end. Nevertheless the standards of doxastic acceptability we embrace in practical contexts when we are making judgements in a careful and reflective manner are such that this testimony needs to overcome an exceptionally high hurdle. Unless the plausibility that this testimony is mistaken or deliberately deceitful is even lower than the extremely low plausibility that attaches to the supposition that a pervasive and welltested regularity that has previously manifested itself throughout all human history has been breached at a particular time and place, it is not appropriate for us to accept that this testimony is correct. And although testimony of this quality is at least conceivable, Pritchard and Richmond hold that even when we assess Hume’s arguments from the perspective of Bayesian reasoning or the non-reductionist view that testimony can possess some independent credibility that does not ultimately derive from nontestimonial sources, it is apparent that Hume manages to present a strong case for the conclusion that such exemplary testimony has never yet been forthcoming in the case of any allegedly miraculous event presented 10

9780826443595_Intro_Final_txt_print.indd 10

2/9/2012 9:37:58 PM

INTRODUCTION O’Connor emphasizes the role Hume allocates in his causal story to the human disposition to anthropomorphize phenomena. People living in relatively sophisticated societies, where information about the genuine hidden causes of otherwise puzzling phenomena is fairly widespread, often engage in what we might term playful anthropomorphism as an amusement or a deliberately chosen form of metaphor. O’Connor argues that Hume contrasts such playful anthropomorphism with a literal-minded anthropomorphism that characteristically emerges when people have little or no grasp of the true aetiology of striking or potentially dangerous phenomena. In those circumstances the human proclivity to think in terms of otherwise unexplained phenomena as arising from agency and intention exerts itself with full force; and as no observable agents can be detected with the appropriate intentions and purposes, the idea develops of multiple invisible and intelligent powers that have a concern with human affairs. Such speculative notions are, of course, theoretically distinguishable from actual beliefs. But O’Connor locates the mechanism that takes people from a spontaneous conception of invisible and intelligent powers to belief in the existence of such powers in human fear and an acutely distressing sense of vulnerability. Once those lively and pervasive passions are engaged, a mere picture of the world is transformed into a set of beliefs that guide people’s actions. In particular, people attempt to relieve their helplessness and sense of vulnerability by treating these hidden agents as susceptible to manipulation by flattery and supplication. Ironically many theists would probably be happy to endorse this or some similar account of the origins of polytheism. O’Connor, however, argues that Hume’s account of the psychological genesis of theistic religion is every

true than some specified rival hypothesis. He accepts that Hume is indeed inclined to maintain that no detailed hypothesis that we can formulate about the ultimate origins of the universe and the order it displays is worthy of endorsement as an explanation that is more likely to be true than false. But he maintains that Hume takes the view that the empirical evidence, inadequate as it is in terms of favouring a particular determinate theory as the most likely theory, does at least marginally favour a naturalistic account of the universe as more plausible than the world-view represented by theism and conventional deism. As Hume was only too well aware, the apparent paucity of good empirical evidence for theism or conventional deism has not prevented the emergence of popular theistic religions with vast numbers of nominal adherents. How can the widespread prevalence of this form of belief be explained if it is not a response to evidence; and even if theistic religions potentially lack the virtue of offering a true description of the universe and our place within it, could it be the case that the existence of such religions is a vital bulwark of morality and an expression of the highest and most sublime aspects of human nature? David O’Connor investigates the account Hume provides of the psychological origins of religious belief, and he contends that this account is a strongly deflationary one. Not only does Hume describe in the Natural History of Religion a set of psychological mechanisms that explain how religions can arise and sustain themselves even if their core metaphysical and historical claims are both unwarranted and untrue, but he also presents some of those mechanisms as dependent on such unedifying aspects of human nature as ignorance, fear and servile self-abasement. 11

9780826443595_Intro_Final_txt_print.indd 11

2/9/2012 9:37:58 PM

INTRODUCTION those affected by that action, we are drawn to feel a certain moral sentiment of approbation. The feeling of such a moral sentiment constitutes, for Hume, a moral judgement. Given that moral judgements depend on our emotional responses to others, they must be, at least in some sense, subjective since they are not independent of our natural human responses to each other. Driver, however, explores how moral judgements can nevertheless possess a kind of objectivity in that moral truths are independent of an individual’s particular responses to a certain action. Such objectivity is supplied by our ability to adopt the general point of view. We can ‘correct’ our sometimes misguided moral judgements because we are able to adopt a perspective divorced from our own, a perspective encompassing the ‘narrow circle’ of those affected by a certain action and not biased by our own concerns or interests. Can, then, a moral judgement be true or false? Driver argues that it can, the ultimate grounding for the truth of a moral judgement lying in the utility of the actions that we judge to be virtuous from the general point of view. Dan O’Brien continues to explore Hume’s account of morality, focusing on his conception of virtue and vice. Virtues for Hume are those character traits that are useful and agreeable to ourselves and to others, and thus people manifest many different kinds of virtue and many different vices. Hume denies that all virtues are innate and God-given and highlights the importance of artificial virtues, traits that people in societies have developed in order to aid the social interactions within those communities. We come to be able to see certain traits as virtuous through sympathizing with the effects that a person’s behaviour has on those around them. Benevolence is virtuous because I resonate to the pleasure

bit as subversive as his explanation of polytheism. Hume, in O’Connor’s judgement, rejects as wholly inconclusive the supposed evidence for theism from miracle reports and such arguments as the cosmological argument and the design argument. And the emergence of theism from a background of antecedent polytheism is explained by Hume in terms of an attempt by people to ingratiate themselves with a particular invisible power by assigning to that agent ever more impressive attributes and abilities in much the same way that one might seek to curry favour with a murderous human despot by eulogizing his or her non-existent qualities of wisdom, justice and benevolence. This practice of base flattery inevitably corrupts over time even the judgement of the original flatterers, and Hume sees its impact on the beliefs of other people in society, especially when aided by education and religious instruction, as even more profound and pernicious. In this manner, some previously negligible deity of comparable status to a petty human princeling is potentially exalted over many generations, without any assistance from cogent truthoriented reasoning, into the supposedly omnipotent, infinitely perfect and wholly just creator of the entire universe. Of a piece with his attitude towards, and arguments against, religion Hume provides a secular moral theory, one in which there is no place for God. Julia Driver turns to this account and its grounding in the natural emotional responses we have to the happiness and suffering of those around us. As O’Brien and Hardin also go on to discuss in subsequent chapters, such emotional responses have their source in our sympathy with others – sympathy, for Hume, seen as the capacity we have to share the emotions of our fellows. And, when the actions of a person lead us to feel pleasure, via sympathy with 12

9780826443595_Intro_Final_txt_print.indd 12

2/9/2012 9:37:58 PM

INTRODUCTION behaviour that serves the interests of other people is primarily a product of psychological mirroring and the conventions that arise when self-interested agents of limited power are attempting to maximize their own benefits from repeated interactions with other similarly self-interested agents. Hardin describes psychological mirroring as an automatic response that people show to the actions and emotional states of other people. It is readily observable that human beings have a strong tendency to mimic the behaviour of the people around them. However, it also seems to be the case that most of us find other people’s observed emotional states similarly infectious. Observing a person showing clear signs of distress or fear tends to give rise to analogous emotional states in the spectator. And behaviour manifesting joy and gladness has at least some tendency to raise the spirits of a person who observes such behaviour. Hardin credits Hume with being one of the first thinkers to explore in any detail the implications of this phenomenon for human actions and choices. Given the existence of psychological mirroring, the psychological states of other people cannot be a matter of complete indifference to us. No matter how self-interested we happen to be, our own lives are more satisfactory, all other things being equal, when the people around us are also faring well. And this responsiveness to the psychological states of other people is what Hardin identifies as lying at the core of the Humean principle of sympathy. Sympathy alone, however, is an inadequate explanation for the range of circumstances in which people seem to accept some check on their self-interest so that the interests of other people can be safeguarded or promoted. Hardin accordingly attaches great importance to Hume’s exploration of the way in which

that the benevolent person’s actions bring to his friends and acquaintances. Reason, however, also plays a role here, but only in helping us to appreciate what sentiments we should feel if we are to be impartial in the requisite way. O’Brien goes on to explore how Hume subverts the religious conception of virtue or what Hume calls the ‘monkish virtues’; elevating pride to his first natural virtue, a due sense of pride being agreeable to ourselves and ultimately useful in our social engagements. That many traits are useful and agreeable is uncontroversial, but a distinction is also usually drawn between traits that have a moral dimension and those that do not: benevolence and compassion are of the former kind; dexterity and wit, the latter. Hume, however, thinks that distinctions hereabouts are not at all sharp and that moral virtues are not different in kind from other beneficial ways of behaving. Russell Hardin moves the discussion away from questions concerning the nature of moral judgement and what constitutes a good character to a consideration of Hume’s analysis of how self-interest can give rise to social conventions and organizations that promote public benefits. Hardin views Hume as deliberately eschewing the attempt to show that moral principles are true or rationally justified in favour of a scientific investigation into how beings who are predominantly motivated by self-interest nevertheless create institutions and social practices that serve the collective good. Numerous philosophers have implausibly purported to show that altruistic behaviour is a fundamental dictate of reason or a requirement imposed upon us by some divine lawgiver, but Hume is seen by Hardin as adopting the radically different and substantially more illuminating approach of explicating how 13

9780826443595_Intro_Final_txt_print.indd 13

2/9/2012 9:37:58 PM

INTRODUCTION women throughout his life, relationships often based on mutual respect and shared intellectual interests rather than transient sexual or romantic passion – although it is also clear from her account that such passion was certainly not wholly alien to Hume’s character. Even more importantly Guimarães identifies Hume’s writings as showing a great willingness to deconstruct gender dichotomies. In his History of England, female characters are frequently portrayed as active agents endowed with energies, drives and reasoning abilities that are at least equivalent to anything possessed by the men surrounding them. And Guimarães argues that when Hume is explicitly engaged in the study of human nature at a more theoretical level, his emphasis on human beings as embodied mammalian animals responding to the influence of concrete conditions including social circumstances and personal relationships means that he avoids the trap of constructing an idealized account of our nature that uncritically sees its essence as lying in such allegedly masculine virtues as pure rationality and the suppression of the passions. Moreover, Hume’s account of human reason and inference further subverts traditional gender categories by presenting such reasoning as founded in associations of ideas, our passions and the faculty of sympathy. Guimarães also notes that when Hume is working within traditional gendered categories, he usually speaks in favour of the wider diffusion of supposedly feminine characteristics. Such virtues as tenderness, benevolence and mildness are not simply seen by Hume as appropriate for women. Instead he argues that society, as a whole, would benefit from these virtues being more widespread amongst men as well. Hume readily acknowledges that the martial virtues of aggression, fierceness and intransigence have

repeated interaction between people comes to shape social behaviour in ways that see collective benefits emerging from self-interest. Even a purely self-interested agent needs to return favours if he or she is to have much chance of securing the co-operation in the future of people who are aware of past performances. And if we advance to a more sophisticated level, justice, in the sense of social stability and good order, is something that we all have some interest in promoting even if the institutions and habits that maintain stability and order sometimes prove inconvenient to us on particular occasions. Hardin accordingly presents Hume as someone who succeeds in setting before us a detailed account of the self-interested strategies that lead to the evolution of some of the most salient social practices and forms of organization that serve to enhance our collective well-being. One important aspect of any human society is the relationship between the sexes, and Hume’s views on this relationship and on gendered differences have aroused considerable controversy. Some commentators have accused him of acquiescing in and even actively seeking to defend sexist forms of discrimination and oppression. Other readers of his writings have, in contrast, seen him as someone who seeks to enhance the status of women and also wishes to see some of the values and psychological characteristics traditionally associated with women disseminated more widely throughout society. Lívia Guimarães mounts a strong defence of the view as elsewhere there should be an accent on one of the i’s of Livia that it is the latter reading of Hume’s position that offers the more illuminating perspective on his attitude towards women and conventional distinctions between masculine and feminine characteristics. She helpfully reminds us of Hume’s many close relationships with 14

9780826443595_Intro_Final_txt_print.indd 14

2/9/2012 9:37:59 PM

INTRODUCTION our attention to the care that Hume took to investigate this phenomenon by seeking out the best available data and using his extensive acquaintance with classical authors to compare Europe with the civilizations of the ancient world. Moreover, Schabas maintains that Hume was right to explain this rise in wealth by invoking the combined impact of division of labour and the increased supply of silver coinage made possible by the mines of the New World. Hume also emerges as equally astute in his reflections on the consequences of such additional wealth for human happiness and welfare. Unlike conservative critics of wealth and luxury who saw and often still affect to see such things as harbingers of moral decay, Hume held that the modern commercial world and the opportunities that it generated had an improving effect on civil society and people’s characters. Schabas presents him as arguing that trade and manufacturing promoted civility, gave a new impetus to learning and human ingenuity, and enhanced liberty and equality. Hume saw these benign influences as most readily impinging on those located in the middle ranks of society, but a flourishing middle class helped to bind all of society together in ways that ultimately benefited everyone. So although Hume was fully prepared to disparage the rapacious acquisition of expensive trinkets, Schabas locates in his writings an ingenious account of how wealth indirectly promotes human happiness. Hume’s predisposition towards Stoic values meant that he viewed material possessions in themselves as being of little consequence once the necessities of life had been supplied, but the process of acquiring wealth through trade and participation in manufacturing served the crucial role of giving people the opportunity to gain personal satisfaction and a sense of purpose from the exertion

utility in primitive societies, where violence is needed to maintain order and to repel invasion and despotic oppression. But in more civilized societies Hume sees other virtues as more effective at promoting the general well-being, and Guimarães draws our attention to the fact that Hume frequently indicates that these less abrasive virtues are best spread throughout society by increasing the opportunity for women to exert their influence on men. Guimarães therefore concludes her chapter with the striking suggestion that Hume can be seen both as sketching out an ideal society that would constitute a feminist utopia and as recommending a greater emphasis on supposedly feminine virtues and attributes as an effective means of improving existing societies. Margaret Schabas investigates Hume’s economic thought. Schabas points out that Hume differs from most present-day economists by emphasizing the greater value to the individual of mental well-being rather than material wealth. Nevertheless Hume’s interest in all aspects of human nature and the social forces that shape people’s lives meant that he corresponded and published quite extensively on matters of economic policy and theory. Indeed Schabas argues that the circulation and influence in the eighteenth century of Hume’s economic essays entitles him to be seen as one of the foremost economists of his era. His pre-eminence amongst British theorists was eventually to be usurped by his friend Adam Smith. Schabas reminds us, however, that this was not to happen until the 1790s, despite the publication of Smith’s The Wealth of Nations in 1776 and Hume’s death in that same year. Schabas concentrates in her chapter on Hume’s response to the conspicuous rise in the wealth of Western Europe in the seventeenth and eighteenth centuries. She draws 15

9780826443595_Intro_Final_txt_print.indd 15

2/9/2012 9:37:59 PM

INTRODUCTION as such a judge. The thought here is that aesthetic pronouncements issuing from people who lack those attributes can be dismissed in much the same way as the colour judgements of someone known to be suffering from a fever or viewing an object in non-standard lighting conditions carry no weight with us in respect of our assessments of the object’s real colour. This, however, raises the question of whose judgement is to be accepted if and when people who count as qualified judges disagree. In some specific instances Hume is perfectly content to say that the disagreement is irresoluble. But Rowe points out that Hume is not always so accommodating: sometimes he seeks a standard of taste that can override or correct the judgement of qualified judges. Rowe maintains that all Hume’s attempts at explaining how such corrections can be given legitimacy are unsuccessful. Hume sometimes appeals to rules of composition, but Rowe powerfully argues that these rules, on a Humean account, can be nothing more than inductive generalizations that summarize the characteristics displayed by works that people usually find pleasing and beautiful. Thus they lack the normative force that Hume requires. Similarly, Rowe rejects Hume’s alternative appeal to a consensus amongst qualified judges. Even if majority opinion were against your personal verdict, would it make sense for you, as a person with the attributes requisite for being a qualified judge, to treat your judgements as wrong simply because you are in a minority? Finding yourself in a minority might well give you grounds for reviewing your reactions to a particular work or artistic performance again. But being in a minority of qualified judges is not constitutive of being wrong in your aesthetic judgement. Rowe concludes that although Hume rightly sees the need to

of their mental and physical powers in an undertaking that was immediately appealing to them. As might be expected with Hume, this confidence in the ameliorative powers of the commercial world is hedged around with substantial reservations. Schabas indicates that these reservations presciently included worries about the destructive power of public debt in the hands of politicians and the prediction that the American colonies and China would eventually eclipse Britain and other European nations in terms of trade and manufacturing output. But Schabas amply succeeds in showing that Hume’s case for supposing that commerce and the pursuit of wealth can often improve people’s dispositions and moral character remains a useful antidote to the unthinking prejudice that morality and personal development are best promoted by austerity and the eschewal of luxury. Mark Rowe, in contrast, is less sanguine about the merits of Hume’s account of a standard of taste in matters of aesthetic judgement. Rowe argues that Hume is concerned to reconcile a form of subjectivism about aesthetic taste with the supposition that some aesthetic judgements can be mistaken. The subjectivism is a product of Hume’s commitment to the view that aesthetic qualities are projected onto the world rather than discovered in the world. But the need to find some room for the concept of an error in aesthetic judgement arises from Hume’s conviction that some judgements of aesthetic merit would be as perverse and as obviously illegitimate as some patently false pronouncements about physical objects and such qualities as size and shape. A part of the answer to Hume’s dilemma lies in the concept of a qualified judge, and Rowe enumerates the attributes that Hume regards as essential if someone is to be viewed 16

9780826443595_Intro_Final_txt_print.indd 16

2/9/2012 9:37:59 PM

INTRODUCTION conclusions supported by real evidence. At the same time the writer requires many of the skills of the poet or dramatist: the truth must be shaped and ordered so that a reader readily enters into the narrative and engages with the character and situation of the personages presented to him. Costelloe emphasizes that Hume sees history written in this manner as serving the crucially important function of laying out the past before us so that it can serve as a guide to future conduct. Firstly, it allows distant past events to be used by the scientist of human nature as a means of confirming or refuting hypotheses about our mental mechanisms and dispositions. Thus it provides the philosopher, in his role as a psychological anatomist, with the data he needs to guide and refine his conclusions. However, it also serves the second function of improving our moral judgements. People and events close to us in time and space are frequently assessed through a prism of partiality that prevents us from seeing how they strike other people, and important potential consequences have often not yet had a chance to manifest themselves. In contrast, if we are reading a historical narration of events that took place many years ago that involved people not intimately connected to us, we have an opportunity to arrive at less biased and better-informed moral judgements, a habit that can then be transferred to situations in which our own interests are at stake. Thus Costelloe maintains that Hume regarded his philosophical and historical investigations as seamlessly intertwined. Just as abstruse philosophy is depicted in the first section of the Enquiry concerning Human Understanding as guiding us to a better understanding of human nature and easy philosophy is portrayed as inspiring us to act virtuously, so too ‘philosophical’ history continues the task of laying

accommodate critical discussion when considering questions of aesthetic merit, his view of what constitutes argument in that arena is too impoverished to explain how genuine debate can take place and real discoveries can be made. Timothy Costelloe introduces us to Hume’s conception of ‘philosophical’ or ‘true’ history, and his analysis of what Hume means by such history helps to bind together Hume’s more explicitly philosophical works and his History of England. Some of Hume’s more malicious critics have accused him of effectively abandoning philosophy after the poor reception afforded the Treatise in order to pursue money and popular fame through the alternative means of writing a best-selling history. However, Costelloe brings into focus numerous important continuities between the philosophical project pursued in the Treatise and Hume’s aspirations for his History. Costelloe argues that Hume sees ‘philosophical’ history as an attempt to combine responsiveness to evidence, rather than the promptings of partiality or the imagination, with a reconstruction of the past that sustains the reader’s interest and gives him or her a lively sense of the truth of the events portrayed. A mere propagandist concentrates only on the second of these two tasks; but if the author makes no attempt to select events and shape the narrative in a way that will appeal to the reader’s imagination and powers of empathy, then the resulting work will be entirely unreadable. Thus a writer of the kind of history that Hume views as worthy of a philosophical author needs to have the skills to weigh testimony carefully, a passion for the pursuit of the truth that promotes impartiality and overrides any temptation to flatter influential patrons, and the capacity to keep the imagination in check so that fanciful associative links do not crowd out 17

9780826443595_Intro_Final_txt_print.indd 17

2/9/2012 9:37:59 PM

INTRODUCTION pattern of interpretation that stresses his radical empiricism and his affinities with Locke and Berkeley. Critics of empiricism have seen Hume as providing the valuable service of exposing how radical empiricism ultimately collapses into incoherence, whereas philosophers of empiricist sympathies have often purported to find in Hume the inspiration for developing what they hoped would be a successful version of empiricism that eschewed all a priori metaphysical speculations. Russell is not inclined to deny that this interpretative tradition has inspired much important philosophical work, but he does deny that it offers the resources required to construct an accurate interpretation of the underlying nature of Hume’s philosophical stance. The difficulty that Russell identifies is that it has become increasingly clear that the empiricist elements of Hume’s thought co-exist with a naturalistic programme that involves the construction of an intricate science of human nature that purports to be based on experience and experiment. However, the interpretation of Hume as a radical empiricist seems to have sceptical implications that are inconsistent with the development of such a science. Yet if we water down the empiricism in order to make it more compatible with the positive side of Hume’s philosophy, it remains the case that Hume’s philosophical writings appear to contain an array of explicitly sceptical arguments that do not need to be embedded within a framework of radical empiricism in order to pose a serious challenge to his naturalistic project. Russell strikingly sums up the situation as generating the worry that Hume’s philosophical outlook is ultimately broken-backed. The sceptical aspect of his philosophy, which seems to be clearly present even if it is not to be construed as generated by a radical form

bare the hidden mechanisms of human action while simultaneously depicting virtuous and vicious characters in such a light that we sympathize with the former and are repelled by the latter. Paul Russell brings the volume to a conclusion with a discussion of changing trends in the interpretation of Hume’s philosophical position. Russell distinguishes between the interpretation of a philosopher’s position and the legacy of that position. Interpretation is a matter of arriving at an understanding of a philosopher’s original aims and intentions, whereas the legacy is constituted by the reception of his or her views and their fruitfulness over time. The various competing interpretations that might arise form part of that reception, but there need be no correlation between the dynamism of the interpretative framework and its philosophical fecundity. Similarly, a sensitive and well-balanced interpretation might reveal itself over time to be nothing more than the accurate signposting of a barren philosophical cul-de-sac, whereas an interpretation that is little more than a caricature of a philosopher’s actual project might fortuitously inspire subsequent philosophical developments of great value and independent interest. Russell, however, warns against the error of supposing that an important and interesting legacy confirms the accuracy of the interpretation that generated it. He also points out that if we complacently allow an interpretation of a philosopher’s views to go unchallenged because it is linked to a valuable legacy, we are in danger of forgoing important philosophical developments that might arise from reflection on some plausible alternative interpretation. In the specific case of Hume, Russell argues that much of the fruitfulness of his legacy up to this point has arisen from a 18

9780826443595_Intro_Final_txt_print.indd 18

2/9/2012 9:37:59 PM

INTRODUCTION arguments with the more positive aspects of his philosophical position. Whether Russell is right to imply that this is both necessary and sufficient to permit such a reconciliation is not yet clear. However, other work by Russell has certainly undermined the supposition that the contents of the Treatise lack a substantial connection to issues of religion. And it can safely be asserted that solving the puzzle of how to harmonize the sceptical and positive sides of Hume’s philosophy is now widely acknowledged to be one of the principal tasks that needs to be accomplished if we are ever to possess a truly satisfactory interpretation of all the essential elements of Hume’s philosophical perspective.

of empiricism, does not initially appear to cohere well with Hume’s science of human nature. Russell accordingly maintains that the way forward is to place Hume’s philosophical writings in a new interpretative framework, one that sees Hume not as part of a triumvirate of British Empiricists or a follower of Newton or Hutcheson but as someone who is actively attacking the metaphysical and moral foundations of Christianity as a member of a partially concealed tradition of ‘speculative’ atheism. In Russell’s judgement, recognition of Hume’s irreligious intentions as manifested even within the Treatise provides the key to an account of Hume’s writings that can reconcile his sceptical

19

9780826443595_Intro_Final_txt_print.indd 19

2/9/2012 9:37:59 PM

1 HUME’S LIFE, INTELLECTUAL CONTEXT AND RECEPTION Emilio Mazza

in . . . understanding, dignity of mind, would bereave even a very good-natured, honest man of this honourable appellation. Who did ever say, except by way of irony, that such a one was a man of great virtue, but an egregious blockhead?’ (EPM App. 4.2 / 314).4 Young Hume was troubled by a ‘weakness’ of spirits; later on he would see a significant relationship between ‘delicacy’ and ‘weakness’ of the mind (LDH 1.17, 3; 1.397, 214). The first as well as the last edition of his Essays open with ‘Of delicacy of taste and passion’, and only in 1772 does Hume stop claiming a ‘very considerable connexion’ between these delicacies in the original frame of the mind (E 603). His mother’s supposed saying has been discussed for 150 years by those who seek to defend the reputation of ‘one of the greatest philosophers of any age, and the best friend to mankind’, as d’Holbach calls Hume, without contradicting a woman of ‘singular merit’, as Hume calls his mother.5 At the end of his Life Hume celebrates himself as ‘a man of mild dispositions, of command of temper, of an open, social, and cheerful humour’, and his temper as ‘naturally cheerful and sanguine’ and not ‘very irascible’ (MOL, LDH 1.1–3). In

These are a few particulars, which may perhaps appear trifling, but to me no particulars seem trifling that relate to so great a man (W. Cullen to J. Hunter, 17 September 1776)1 1. ‘WAKE-MINDED’ ‘Our Davie’s a fine good-natured crater, but uncommon wake-minded’, Hume’s mother is supposed to have said in a piece of familial assessment.2 And with regard to Hume’s religious principles, his brother John ventured the opinion: ‘My brother Davie is a good enough sort of man, but rather narrow minded’.3 This latter judgement echoes in its choice of words Hume’s own recollection that in Paris they ‘used to laugh at me for my narrow way of thinking in these particulars’ (LDH 2.273, 484), and his description of Rousseau – ‘a very agreeable, amiable Man; but a great Humourist’ (LDH 2.13, 303; see LDH 2.130, 381) – indicates that Hume shared with his brother a partiality for verbal sallies that combined initial restrained praise with a less commendatory ending. We seem too to find in the Enquiry concerning the Principles of Morals, a riposte on Hume’s part to his mother’s assessment of his character: ‘any remarkable defect 20

9780826443595_Ch01_Final_txt_print.indd 20

2/9/2012 9:39:42 PM

HUME’S LIFE are the most ‘unfortunate’ books, and the Political Discourses a work ‘successful on the first publication’ and ‘well received abroad and at home’. The more sustained success begins in the 1750s, when Hume discovers symptoms of a ‘rising reputation’, including, for example, a ‘railing’ reaction of the clergy (MOL, LDH 1.2–4).

1757, somewhat between jest and earnest, Hume says he is a ‘good-natured man of a bad character’ (LDH 1.264, 139) and also (if the text be Hume’s) a ‘very good man, the constant purpose of whose life is to do mischief’.6 The history of his writings shows him as a man of ‘superior genius’, a quality that he does not even recognize in d’Alembert, who was simply a man of ‘superior parts’, even though after Paris Hume considered him ‘with some few exceptions (for there must always be some exceptions) . . . a better model of a virtuous and philosophical character’ (LDH 2.110, 363).

3. ‘NEVER TO REPLY TO ANY BODY’ In about 20 years (1739–61) Hume publishes almost every kind of writing: a Treatise, its Appendix and Abstract; a Letter to a friend and a True Account of the conduct of another friend, the Essays, the Philosophical Essays and the Enquiries; the Discourses and the Dissertations; the History, the Natural History and a Dialogue. He also receives almost every kind of answer. In 1766 he observes: ‘I could cover the Floor of a large Room with Books and Pamphlets wrote against me, to none of which I ever made the least Reply, . . . from my Desire of Ease and Tranquillity’ (LDH 2.92, 351). With regard to the years 1749–51, in 1776 Hume declares he has ‘fixed’ and ‘inflexibly maintained’ a resolution ‘never to reply to any body’ (MOL, LDH 1.3). He starts asserting this resolution in the second half of the 1750s, as a reaction to the ‘Warburtonian School’, but in 1760 he declares that he formed it ‘in the beginning of my Life, that is, of my literary Life’ (LDH 1.320, 172), which seems therefore to begin with a commitment to refuse any literary controversy. Like many official claims, however, this is not completely reliable even though it does contain a substantial admixture of truth. He often replies indirectly to his critics in his writings, and sometimes he is even driven to the expedient

2. MY OWN (UNSUCCESSFUL) WRITINGS In 1734, when he begins ‘to despair of ever recovering’ from his ‘Disease of the Learned’, Hume wrote ‘a kind of History of my Life’ (LDH 1.13, 3; 1.17, 3); in 1776, when his life is really ‘despaired of’, he writes ‘the History of my own Life’ or My own Life (LDH 2.318, 522A; LDH 2.323, 525). This short ‘funeral oration of myself’, Hume says, contains ‘little more than the History of my Writings’, since ‘almost all my life has been spent in literary pursuits’ (MOL, LDH 1.7). The rhythm of the Life is the alternation of learning and business, expectations and disappointments, which recalls that of action and repose in the ‘Refinement in the Arts’ (E 270). Every disappointment is overcome by character, ‘command of temper’ and ‘cheerful humour’ (MOL, LDH 1.7). Hume’s Life is also a history of the reception of his writings, where he commonly distinguishes between immediate and gradual success. The want of it is chiefly measured by the standard of silence. The Treatise and the first volume of the History 21

9780826443595_Ch01_Final_txt_print.indd 21

2/9/2012 9:39:43 PM

HUME’S LIFE both in the fact that he has never ‘preferred a request to one great Man, or even . . . [made] advances of friendship to any of them’ and in the fact that he has nevertheless found himself on good terms with such people in his personal affairs, public business, and while composing his History (MOL, LDH 1.5–6; LDH 1.113, 63; 1.295, 156; 1.355, 191; 1.427–28, 232; 2.188, 422). As a man now beyond middle age working for the Northern Department, he finds that ‘to a Man of a literary turn, who has no great undertaking in view, . . . public Business is the best Ressource of his declining Years. Learning requires the Ardor of Youth’ (LDH 2.385, 137). Thirtythree years before, in spring 1734, trying to leave his distemper behind and working on the Treatise, he found ‘two things very good, Business & Diversion’, and resolved ‘to seek out a more active life’, laying ‘aside for some time’ his pretensions in learning (LDH 1.17, 3; MOL, LDH 1.1). In 1767 the Earl of Rochford remembers that Hume is ‘unfit for business’, and Hume himself has already admitted that the office of secretary requires ‘a Talent for speaking in public to which I was never accustomd’ (LDH 1.519, 289).9 However, Hume’s essay ‘Of Eloquence’ (1742) attacks, following Swift and La Bruyère the ‘antient Prejudice industriously propagated by the Dunces in all Countries, That a Man of Genius is unfit for Business’ (E 621), and the first Enquiry (1748) claims that the accuracy of abstruse philosophy is ‘subservient’ to every art or profession: the politician, the lawyer and the army general may take advantage from it (EHU 1.9 / 10). In part, at least, this sounds like a defence of those aspects of his life and career that were not directly connected to his literary and philosophical pursuits, for at various times he found himself taking on the roles of clerk for a merchant in Bristol

of explaining that he is giving an answer and that it should be extended to different adversaries, as in the 1775 ‘Advertisement’ to the Enquiries (LDH 2.301, 509). In 1757 someone suggests that he has deliberately ‘so larded his Work with Irreligion’ that the first volume of the History ‘might sell’,7 and Hume observes that the few ‘Strokes of Irreligion’ are of ‘small Importance’, even though they are likely ‘to encrease the sale’ (LDH 1.250, 132; 1.256, 136). A few months later he allows that he would accept the challenge to defend The Natural History of Religion against Warburton’s criticisms were he attacking his ‘principal Topics’. As he tells the bookseller: ‘The Hopes of getting an Answer, might probably engage [Warburton] to give us something farther of the same kind; which at least saves you the Expence of advertising’ (LDH 1.265–67, 140). Concerning his no-reply resolution, in 1758 Hume still maintains that he ‘shall probably uphold it to the End of [his] life’. The Concise and Genuine Account of his dispute with Rousseau recalls that Hume ‘hath seen his writings frequently censured with bitterness . . . without ever giving an answer to his adversaries’, yet, in the case of the Rousseau imbroglio, the ‘circumstances’ were such as to draw Hume into a scandal, ‘in spite of his inclinations’. Consistent with them, he authorizes the editors to declare ‘that he will never take the pen again on the subject’.8

4. NOT UNFIT FOR BUSINESS: ‘THE ARMY IS TOO LATE’ Like Lucian in De mercede conductis and Apologia pro mercede conductis, which Hume first quotes in 1751–2, he takes satisfaction 22

9780826443595_Ch01_Final_txt_print.indd 22

2/9/2012 9:39:43 PM

HUME’S LIFE dissipation; yet always returned to my closet with pleasure’ (LDH 1.451, 244). In 1746 Hume receives an ‘unexpected’ invitation from St Clair to go with him as secretary in his military expedition, which was planned to be an attack on French Canada but came to its conclusion on the coast of Brittany (MOL, LDH 1.2; LDH 1.382, 206; 1.92, 51; NLH 24, 10). He arranges his ‘Departure for America’ (‘Such a Romantic Adventure, & such a Hurry’) with one box of books and one of paper in his trunk (LDH 1.90, 50).10 Being asked whether he ‘would incline to enter the Service’, he answers that at his years he could not ‘accept of a lower commission than a company’ (LDH 1.94, 52). One year after he says that for the ‘Army [it] is too late’ (NLH 26, 10). The expedition is a ‘failure’, but it gives rise on Hume’s part to a beautiful letter to his brother, a brief journal or hypomnema, a piece on ‘The descent on the coast of Brittany in 1746’, and possibly an article (LDH 1.99, 54; 1.94–8, 53). The expedition also shapes Hume’s opinions about soldiers. Major Alexander Forbes, for example, is described as ‘a Man of the greatest Sense, Honour, Modesty, Mildness & Equality of Temper in the World’: ‘His Learning was very great for a man of any Profession, but a Prodigy for a Soldier. His Bravery had been try’d & was unquestion’d.’ When Forbes kills himself as a result of anxiety and fear that he may have been guilty of a dereliction of duty, Hume maintains that in the course of dying from his self-administered injuries, he expressed a ‘steady Contempt of Life’ and ‘determind philosophical Principles’. And after Hume has seen his friend die in front of him, it is probably Hume who also undertakes the official duty of recording that one Dougal Steuart was made Captain ‘in room of Alexr Forbes deceast’ (LDH 1.97, 53).11

(1734), companion and tutor of a marquess near St Albans (1745–6), secretary to a general and judge-advocate in Lorient (1746–7), secretary and aide-de-camp to the same general in Vienna and Turin (1748), librarykeeper in Edinburgh (1752–7), secretary to the Embassy and Chargé des affaires in Paris (1763–6), and Under-Secretary of State for the Northern Department in London (1767–8). He also considers (but ultimately rejects) the ‘not agreeable’ life of the ‘Travelling Tutor’ (LDH 1.18, 3; 1.17–8, 24; 1.35–6, 14; 1. 57–8, 24; NLH 26, 10), even though he is often ‘mortally sick at sea’ (LDH 2.206, 432; 1.214, 105; also see LDH 1.114, 64; 2.95, 352) and claims that ‘Shortness . . . is almost the only agreeable Circumstance that can be in a Voyage’ (LDH 1.105, 56). Every time he is enjoying his solitude Hume receives an ‘invitation’ he cannot refuse (MOL, LDH 1.2, 4, 5–6). According to the correspondence, his life is a permanent yearning for (philosophical) retreat and leisure, continuously thwarted by external circumstances, leading him into some practical business: ‘I lived several years happy with my brother at Ninewells, and had not his marriage changed a little the state of the family, I believe I should have lived and died there,’ he says in 1759 about his own ‘reluctance to change places’, even though in 1763 he has ‘so often changed’ his places of abode that he comes to think that ‘as far as regards happiness, there is no great difference among them’ (LDH 1.295, 156; 1.415, 224; see also LDH 1.243, 128; 1.246, 130; 1.531, 295; 2.189, 423). With regard to the years spent with General St Clair, the Life claims that ‘these were almost the only interruptions which my studies have received during the course of my life’ (MOL, LDH 1.2–3), even though 15 years before he has allowed: ‘I have frequently, in the course of my life, met with interruptions, from business and 23

9780826443595_Ch01_Final_txt_print.indd 23

2/9/2012 9:39:43 PM

HUME’S LIFE though soldiers have their exceptions, Hume passed his St Clair years ‘agreeably, and in good company’ (MOL, LDH 1.2-3): reliable officers, learned physicians, whist-players and humorous people who dedicate themselves to the ‘service of the Ladys’. Among priests, on the contrary, ‘gaiety, much less the excesses of pleasure, is not permitted’ (E 201n). He is ‘in the Army’ and he calls it ‘our Family’ (LDH 1.97, 53; 1.132, 64).14 With these ‘friends or confidents’ – he says with Quintilian and Svetonius, or more simply with Voiture – he can be free ‘in seriis et in jocis, – amici omnium horarum’ (in grave and jocular manners, – friends of all hours) (LDH 1.102, 56). In 1747, when St Clair invites Hume to go over to Flanders with him (LDH 1.108–9, 61), he has ‘a great curiosity to see a real campaign’, notwithstanding his fears of the ‘expense’ and looking ridiculous as a result of ‘living in a Camp, without any Character & without any thing to do’ (NLH 23, 9). Nothing could be ‘more useful’ to his ‘historical projects’. Hume looks forward to picking up a great ‘military knowledge’, by ‘living in the General’s family, and being introduced frequently to the Duke’s’ (ibid.). In 1748 he attends St Clair in his mission to Vienna and Turin, notwithstanding an ‘infinite regret’ for leaving ‘stores of study & plans of thinking’ (LDH 1.109, 61; 1.111, 62). In accordance with the opinions of Lucian, Bayle, Addison and the Guardian, and following the advice contained in a volume by Polybius, which he keeps in his hand (LDH 1.100, 54), he is looking for ‘an opportunity of seeing Courts & Camps’:

In 1734 Hume had compared the soldier’s courage to the devotee’s devotion (LDH 1.21, 4). In 1748 he publishes ‘Of National Characters’, where a few pages could be entitled ‘The Soldier and the Priest’. It is a double reaction to his academic and military adventures in Edinburgh and Lorient. In 1743 Hume reads Leechman’s sermon on prayer, and sends him some remarks on argument and style, together with 22 small faults that the author does not even take into consideration. The sermon, Hume concludes, unavoidably makes his religious author ‘a rank Atheist’ (NLH 10–14, 6). Despite his youthful claim that ‘there is nothing to be learnt from a Professor, which is not to be met with in Books’,12 in 1744 Hume attempts in vain to become professor at the University of Edinburgh, and declares himself extremely surprised that the ‘accusation of Heresy, Deism, Scepticism, Atheism &c’ is supported by the ‘Authority of Mr Hutcheson & even Mr Leechman’ (LDH 1.58, 24). In 1741 Leechman published another sermon on the character of the priest. ‘Of National Character’ is also an answer to him. Leechman claims we can never clearly ‘unvail’ to mankind their ‘hidden hypocrisy’, nor justly contempt the devout worshippers by calling the outward displays of their inward devotion ‘solemn grimaces, and hypocritical airs’;13 Hume replies that the clergymen ‘promote the spirit of superstition, by a continued grimace and hypocrisy’ and this ‘dissimulation often destroys’ their ‘candor’ (E 200n). Hume denounces their conceited ambition, professional faction and persecuting spirit. In contrast, soldiers are ‘lavish and generous, as well as brave’, ‘candid, honest, and undesigning’. Since ‘company’ is their sphere they can acquire ‘good breeding and an openness of behaviour’ and a ‘considerable share’ of politeness (E 199). Even

this knowledge may even turn to account to me, as a man of letters, which I confess has always been the sole object of my ambition. I have long had an intention, 24

9780826443595_Ch01_Final_txt_print.indd 24

2/9/2012 9:39:43 PM

HUME’S LIFE in my riper years, of composing some History; . . . some greater experience of the Operations of the Field, & the Intrigues of the Cabinet, will be requisite, in order to enable me to speak with judgement upon these subjects (LDH 1.109, 61).

that always, even from your earliest Years, did most easily beset you’ (LDH 1.438, 237; 2.353, App. C. V). In Turin Hume becomes bored and sick. Admiral John Forbes called him ‘the sleeping philosopher’, someone says he was ‘affected by a most violent Fever’, some other that he ‘received Extreme Unction in a dangerous illness’. He hangs around with Lord Charlemont and reads Montesquieu’s Esprit (LDH 1.133, 65).18 Consistent with his Treatise, and in the name of ‘sympathy’, he enjoys the pleasure and beauty of extended, fertile, cultivated plains. He wishes to make ‘a short Tour thro’ some of the chief Cities of Italy’, but apparently the Duke of Newcastle rejects the request.19 He does the accounts (as he did in Bristol) and examines the Sardinian documents in the Commissary’s office.20 He writes St Clair’s official letters and copies them into a letter book. He probably suggests passages for St Clair’s letters (like the observation of the historians that ‘Britain has commonly lost by Treaties what she gain’d by Arms’) and certainly receives suggestions for his future writings: ‘Of the Balance of Power’ discusses the peace of Aix la Chapelle, and the dying Hume is still remembering those inconceivably ‘good terms’ that France had granted to Britain.21 Hume’s experience in Turin resumes that begun in Lorient and prepares the way for his 1760s appointments in Paris and London. General St Clair, Lord Hertford and his brother General Conway all wanted Hume with them. St Clair ‘positively refused to accept of a Secretary from the Ministry’, and Hume goes ‘along with him’; some 15 years later in 1763 Hertford is ‘resolved never to see, or do business with his Secretary, and therefore desired [Hume] should attend him’ (LDH 1.111, 62; 1.421, 228). In March 1767 Hume is ‘deeply immersed in study’, when Hertford surprisingly urged him to ‘accept of

St Clair arrives in Turin on the 8 May, Hume and St Clair’s nephew, Sir Harry Erskine, about eight days later; on 29 November 1748 they all set out.15 The result of the mission is a ‘long epistle’, which he calls a ‘sort of Journal of our Travels’ (LDH 1.114, 64; 1.132, 64). Before the departure Hume optimistically contrasts his situation with that of the ‘severe’ Lord Marchmont, who, ‘entirely employed in the severer studies’, suddenly opens his eyes on a ‘fair nymph’ aged just 16 and marries her in a few days: ‘they say many small fevers prevent a great one. Heaven be praised, that I have always liked the persons & company of the fair sex: For by that means, I hope to escape such ridiculous passions’ (LDH 1.110, 61). Ten days after his arrival he is already ‘troubled’ by an ‘indisposition’ connected with the ‘pretty women’ of Turin. After two months he declares an ‘attachment’ for a Countess of 24.16 The Turin-based Madame Duvernan anticipates the Parisian Madame de Boufflers and their extrovert public reputations stand somewhat in tension with Hume’s claim that he, like Mandeville’s perfect sociable benevolent man,17 took a ‘particular pleasure in the company of modest women’ and had therefore no reason to be ‘displeased with the reception I met with from them’ (MOL, LDH 1.7). In summer 1764 he reminds his reverend friend Jardine that ‘A Man in Vogue will always have something to pretend to with the fair Sex’, and Jardine banters: ‘An inordinate Love of the fair Sex . . . is one of those Sins, 25

9780826443595_Ch01_Final_txt_print.indd 25

2/9/2012 9:39:44 PM

HUME’S LIFE been jailed, he would have risked ignominiously losing his life for having or not having drowned himself.24

the office of Depute-Secretary of State under his brother’. He cannot refuse and sees himself ‘embarked for some time in state affairs’. Yet, he says, ‘I foresaw also that a place was offered me of credit and confidence; that it connected me with General Conway’ (LDH 2.123, 374; also see 1.511, 282). Hume says he feels like a ‘banished man in a strange country’. He is not ‘hurry’d with Business’ and commonly attends on the Secretary ‘from ten to three’. He has ‘no more Business than would be requisite for [his] Amusement’ in London (LDH 2.123, 374; 2.127, 377). Hume is not only employed in ‘cyphering and decyphering’: during his public activities he does not forget his opinions. When the burden of diplomatic work at the embassy in Paris is falling entirely upon him (LDH 511–12, 282–3), and friends start calling him ‘a man of Business’ (LDH 1.421n, 228),22 he saves from prison and death an Englishman who attempted to kill himself in the Seine. Marischal Keith congratulates Hume: ‘you have done many good works in your Ministerial functions, I am sure it was one to save a pour fellow from the gallows, who chose rather to drown than starve’.23 And Diderot has the complete story:

It was the time of extravagant requests, like that of the ‘Apulien Philosopher’ Vincenzo Maria Gaudio (1722–74). In January 1764 he wrote to Hume asking him two questions ‘for the good of human kind’: ‘How many and which physical and moral causes produce the variety and contrariety of opinions among men?’; ‘How to reduce the sum of evils and increase that of goods?’.25 When he is Under-Secretary in London, ‘degenerated into a petty Statesman’ and ‘entirely occupyed in Politics’ (LDH 2.128, 379), he meets another extravagant case: ‘one Giraldi, an Italian Physician’. Giraldi, who is in London and needs protection in Italy, addresses himself to Hume; Hume reporting to Lord Shelburne: [He] seems to me a man of sense and learning, and whose orthodoxy has of consequence been brought under great and I suppose just suspicions. . . . It seems a Cardinal, in his absence, fell in love with his wife, and has taken her into keeping; and on the physician expressing some displeasure at this treatment, his Eminence, who has great credit in the Holy See, has threatend to have him put into the Inquisition . . . . He has addressed himself to me, on the supposition, no doubt, that I woud sympathize with his cause. I conjure therefore your Lordship, if there be any virtue, if there be any praise, if there be anything comely or of good report, to save the poor heretic from the flames . . . his case wou’d puzzle Rhadamanthus himself: as a cuckold, he ought to go to heaven; as a heretic to hell. But, without joking, his case is worthy of compassion; and I recommend it to your Lordship’s humanity.26

They fished him out alive. They brought him to the Grand Châtelet, and the Ambassador had to interpose his authority to prevent them from putting him to death. Some days ago Mr. Hume told us that no political negotiation had been more intriguing than this affaire and that he had been obliged to go twenty times to see the first president before he could make him understand that there was no article, in any of the treaties between France and England, that forbade an Englishman from drowning himself in the Seine under penalty of being hanged. And he added that, if his compatriot had unfortunately 26

9780826443595_Ch01_Final_txt_print.indd 26

2/9/2012 9:39:44 PM

HUME’S LIFE of a large library’. At the beginning of the 1760s ‘the copy-money given [him] by the booksellers, much exceeded any thing formerly known in England’ and Hume is ‘not only independent, but opulent’. In 1766 the Parisian Secretary returned to Edinburgh, ‘not richer, but with much more money, and a much larger income’ than he left it. He was now ‘desirous of trying what superfluity could produce’. In 1769 the London UnderSecretary returned to Edinburgh ‘very opulent’ (he ‘possessed a revenue of 1000 l. a year’) and with the double prospect of long enjoying his ‘ease’ and of seeing the increase of his ‘reputation’ (MOL, LDH 1.1–6). Thanks to Hertford’s family he really was, as he once wrote from Paris, ‘in the high Road to Riches’ and ‘in the high Road to Dignities’ (NLH 78, 38; LDH 1.421, 228).

Some ‘fresh intelligence’ discovers to Hume that Giraldi ‘lives in intimacy with Gemino, no great sign of his orthodoxy’. His project was to retire to the Island of Capri, which Giraldi ‘represents as an earthly paradise’, and – Hume concludes – ‘indeed the only paradise he ever expects to go to’.27

5. MY OWN FORTUNE ‘Money – says the Concise Account – is not universally the chief object with mankind; vanity weighs farther with some men’.28 Not entirely exempt from vanity, Hume never abandons the money that belongs to him ‘of right’, like the quarter salary from the Annandale Estate and the half-pay military pension from the Treasury: after more than 15 years he is still fighting for it. But he is also ready to retract his application at the Advocates’ Library, retain the office and give a friend a bond of annuity for the salary. In 1747 he calls himself ‘a good Oeconomist’ (NLH 26, 10). Riches are valuable ‘at all times, and to all men’ (E 276), and in his short Life he spends some words celebrating his income. He says he was ‘of a good family’, but ‘not rich’. As a younger brother, his ‘fortune’ was ‘very slender’ and therefore unsuitable to his literary plan of life. So he laid down a rule: ‘to make a very rigid frugality supply my deficiency of fortune’. In 1745 his Annandale appointments made a ‘considerable accession to [his] small fortune’; in 1746–8 the St Clair appointments earned him a ‘fortune’ that he calls ‘independent’. He wanted ‘to maintain unimpaired [his] independency’ and he is now ‘master of near a thousand pounds’. In the 1750s the Faculty of Advocates gave him ‘little or no emolument’ but the ‘command

6. STRIKE OUT STERNE: FASHION IN PARIS Hume was in Paris, Reims and La Flèche in the 1730s, Paris in 1748 and Paris again in the 1760s. He constantly saw himself through the French looking-glass: the first philosophical readings and the successful French translations of his writings (in 1761 the Essais Philosophiques earn themselves a place in the Index Librorum Prohibitorum), the Embassy, the Court, the Great Ladies (Madame de Boufflers) and the Philosophes (Rousseau). In 1745 Hume first expresses the slightly melancholy intention of retiring to the South of France (NLH 17, 7). In the Life he remembers living in Paris as a ‘real satisfaction’: ‘I thought once of settling there for life’ (MOL, LDH 1.6). Everyone affects to consider him ‘one of the greatest geniuses in the world’ (LDH 1.410, 223), since in 27

9780826443595_Ch01_Final_txt_print.indd 27

2/9/2012 9:39:44 PM

HUME’S LIFE Sterne unveils Hezekiah-Hertford’s ‘vanity’ and ‘ostentation’. Later on, at Hertford’s table, Sterne had a dispute with Hume (a ‘little pleasant sparring’, he says). In his sermon Sterne had celebrated integrity and miracles, and blamed pride and hypocrisy. At dinner ‘David was disposed to make a little merry with the Parson; and, in return, the Parson was equally disposed to make a little mirth with the Infidel’. Sterne concludes: ‘it is this amiable turn of character, that has given more consequence and force to his scepticism, than all the arguments of his sophistry.’33 At the end of 1765 Sterne publishes his Sermons with a probably less ‘unlucky’ and offensive version of ‘Hezekiah’. He is ready to ‘quarrel’ with Hume by calling him a ‘deist’, if he will not add his name to the ‘most splendid list’ of subscribers. The Sermons came out, but Hume’s name was not in the list. In 1767 Hume recalls the ‘usual extravagance’ of Sterne’s productions (NLH 160, 80), and in the Sentimental Journey Sterne plays with Hume the historian, his ‘excellent heart’ and bad knowledge of French. When Sterne dies, Hume subscribes five guineas for his widow.34 In 1773 Hume detects in Brydone’s Tour through Sicily and Malta ‘some Levities, too much in the Shandean Style’, which he advises the author to ‘obliterate’. He also says that Tristram Shandy is ‘the best Book, that has been writ by any Englishman these thirty Years . . . , bad as it is’ (LDH 2.269, 482). Three years later he first writes and then strikes Sterne’s name out of his Life.

Paris, unlike London, a man of letters ‘meets immediatly with Regard & Attention’ (LDH 1.497, 272). ‘Anglomania was the manner of the place,’ Charlemont observes, and ‘Hume’s Fashion’ was ‘truely rediculous’: ‘no Lady’s Toilet was compleat without Hume’s attendante.’ Walpole is more concise: ‘Mr. Hume is fashion itself.’29 Indeed, he was more celebrated for his name rather than his writings, for his economical, historical and anti-religious writings rather than his philosophical opinions, and for his general opinions instead of his precise arguments. The French mode entailed ‘excessive civilities’ (MOL, LDH 1.6), but what was ‘at first oppressive’ in two months ultimately sat ‘more easy’, especially as he gradually recovered the ‘facility’ of speaking the language (LDH 1.417, 225; 1.414, 224; 1.498–9, 272).30 The Life sums up: ‘Those who have not seen the strange effects of modes, will never imagine the reception I met with at Paris’. And Hume reports, in a remark that he was later to strike out, that ‘Dr Sterne told me, that he saw I was [celebrated in town] in the same manner that he himself had been in London: But he added, that his Vogue lasted only one Winter’ (MOL, LDH 1.6).31 In 1762 Sterne does not worship the French goddesses, but, he says, he has ‘converted many unto Shandeism’. In 1764 he preached a sermon deemed ‘offensive’ (he calls it ‘innocent’) at the Embassy Chapel. Hertford has just furnished the new and ‘magnificent’ Hôtel de Brancas, which gave ‘the subject of conversation to the polite circles of Paris, for a fortnight at least’.32 Sterne preaches on the Book of Kings and Hezekiah, who foolishly showed all the precious things that were in his house; even his wives and concubines, adds the preacher. Behind ‘urbanity or the etiquette of courts’,

7. LIFELONG LUCIAN AND THE IRISH SKYTHS ‘Lucien est votre auteur favori, et . . . je l’aime bien autant que vous’ (‘Lucian is 28

9780826443595_Ch01_Final_txt_print.indd 28

2/9/2012 9:39:44 PM

HUME’S LIFE cyphers was part of Hume’s official duties in both Turin and Paris. Murphy was also the editor of The Select Dialogues of Lucian, first printed in 1744. In 1767 Hume compares him to the ‘Royal philosopher Anacharsis’. Murphy usually calls himself ‘Ô Murraghoo Rex’, Anacharsis is one of Lucian’s dialogues and the name of a character in Scytha sive hospes. In 1765 Hume had refused to go to Ireland with Hertford: the Dubliners and the Londoners did not want the Scottish philosopher to make such a visit. Hertford had prepared him an apartment in the Castle of Dublin, but Hume thought it ‘not worth while’: ‘It is like Stepping out of Light into Darkness to exchange Paris for Dublin’ (LDH, 1.514, 285). In Ireland the philosopher and historian was ‘excessively disliked’. It will be ‘an Age or two at least’ before the Irish can perceive his doctrines, and ‘perhaps an age or two more’ before they can relish them, writes Chaplain Trail: ‘I could almost as soon promise Antichrist himself a welcome Reception.’36 Possibly alluding to Hume’s account of the ‘most barbarous’ cruelty allegedly perpetrated by the Irish during the ‘universal massacre’ of the English in 1641, where ‘[n]o age, no sex, no condition was spared’ (H 5.55, 341), Murphy says in June 1767 that Hume considers the Irish ‘Savages’ because they ‘eat Human Flesh when [they] can get it good’. The native Irish, adds Murphy, are ‘provd by History to be Scythians by Descent, or rather . . . Skyths, which word has been corrupted into Scots’.37 In a swift Lucianic style Murphy invites Hume to Ireland, ensuring him he will be treated ‘as safe, as kindly . . . as ever [he] was in Paris, or Edenburgh’:

your favourite author, and . . . I love him as much as you do’), Morellet reminds Hume in 1766.35 Lucian follows Hume throughout his literary career. In 1742 he allows that ‘some Dialogues’ of Lucian are among the few excellent pieces of pleasantry in ancient literature (E 134). The explosion of Lucian occurs in the second half of the 1740s. ‘The Sceptic’ (1753 version) suggests that we can improve our mental disposition by reading the ‘entertaining moralists’ and engaging with the ‘imagination of Lucian’, if nature has endowed us with a ‘favourable’ temper (E 179n). Moreover, the moral Enquiry assesses Lucian as ‘licentious with regard to pleasure’ but a ‘very moral writer’ in other respects, and accordingly regards it as highly significant that he ‘cannot, sometimes, talk of virtue, so much boasted, without betraying symptoms of spleen and irony’. In Great Britain, adds Lucianic Hume, such a ‘continued ostentation’ of public spirit and benevolence inclines men of the world ‘to discover a sullen incredulity on the head of those moral endowments’ (EPM 6.21 / 242). In the first Enquiry, where he laments the ‘harsh winds of calumny and persecution’ directed against philosophy, Hume bitterly observes: ‘it does not always happen, that every Alexander meets with a Lucian, ready to expose and detect his impostures’ (EHU 11.2 / 132–3; EHU 10.23 / 121). In all antiquity, says the ‘Populousness of ancient nations’, there is not a philosopher ‘less superstitious’ than Lucian (and Cicero). The ‘agreeable’ Lucian, says the Natural History, had ‘employed the whole force of his wit and satire against the national religion’ (E 463n; NHR 12.174). Morellet is not the only translator of Lucian with whom Hume was acquainted. In Turin he met Edward Murphy (1707–77). Murphy’s repeated ‘grand query’ to Hume concerns a cypher he invented, and the use of

We do not devour Strangers who visit us as Friends; not even such as, we know, 29

9780826443595_Ch01_Final_txt_print.indd 29

2/9/2012 9:39:44 PM

HUME’S LIFE come to rob us. If you will not send me a good Answer to my Query, I will go at the Head of my Mighty Men and extirpate your Nation. Yea, we would eat you all up, Man, Woman, and Child; but that we are a little nice about the Quality of our Fleshmeat. I never saw a Piece of a Scotchman (though we have a whole Province stock’d with them) at any genteel Table here.38

‘all [his] manuscripts . . . desiring him to publish [the] Dialogues’. Hume reserves himself the power to alter his will at any time, ‘even in death-bed’.43 And so he does. He first leaves ‘entirely’ to Smith’s discretion ‘at what time . . . , or whether’ to publish the Dialogues (LDH 2.316–18, 522–522A). Smith accordingly thinks he has persuaded Hume to allow him to leave the Dialogues concerning Natural Religion unpublished if he views that as advisable. Hume feels that his own death is imminent, and begins to think in terms of printing a ‘small edition’ and giving his editor Strahan the ‘literary property’ (LDH 2.323, 525).44 In a first codicil he makes Strahan ‘entirely Master’ of his manuscripts (LDH 2.325, 527): the Dialogues must be ‘printed and published any time within two Years after [his] Death’, and the Life ‘prefixed to the first Edition of [his] Works’.45 In a second codicil he ordains: ‘if my Dialogues . . . be not publisht within two Years and a half after my Death, as also the Account of my Life, the Property shall return to my Nephew’.46 Smith criticizes this ‘unnecessary clause’,47 and Hume leaves Smith a ‘security’ copy (LDH 2.334, 538). Finally, two days before dying, he informs Smith and Strahan that he is leaving his nephew the ‘property of the Manuscript in case by any accident [to Strahan’s Life] it should not be published within three years after [his] decease’ (LDH 2.336, 540). Smith is trying to move away from the Dialogues: ‘If you give me leave I will add a few lines to your account of your own life.’ It would make ‘no disagreeable part of the history’ to relate Hume’s ‘want of an excuse to make to Charon, the excuse you at last thought of, and the very bad reception which Charon was likely to give it’. Inspired by a preceding letter to Wedderburn, Smith wants to celebrate Hume’s ‘steady cheerfulness’

The pretended ‘Reverend Murphy’, who in Rome bought a papal plenary indulgence for three crowns and made some remarks on ‘the pope and his fellow jugglers’, denounces Lucian’s ‘entire want of Candour, while he talks against the Christian Religion’, yet, he adds, ‘it is impossible not to admire his matchless Abilities’.39 Murphy’s translation of Lucian seems to have been a (Greek) textbook at Trinity.40 Three months before dying the ageing Hume invites his nephew David to read Lucian ‘sometimes’ and not to forget his Greek, to mix the volumes of ‘taste and imagination’ with ‘more serious reading’, as the young Hume used to read at his pleasure ‘sometimes a Philosopher, sometimes a Poet’.41

8. HUME AND SMITH: A LIVING SUMMER DIALOGUE Lucian’s writings have something to say about Hume’s death and legacy, the publication of Hume’s Life and Dialogues, and the role taken by Adam Smith: an alternation of intentions and second thoughts, trust and worries, will and codicils. In 1773 Smith is not well and leaves Hume ‘all [his] literary papers’ and the publication of the ‘history of the Astronomical Systems’;42 on 4 January 1776 Hume is seriously sick and leaves Smith 30

9780826443595_Ch01_Final_txt_print.indd 30

2/9/2012 9:39:44 PM

HUME’S LIFE towards death.48 In the January will Hume had asked Smith to take the pains of ‘correcting and publishing’ the Dialogues; now Smith offers to publish an Addition to Hume’s Life and to ‘correct the Sheets of the new edition of [his] works’, the Dialogues excepted.49 Hume answers Smith that he is ‘too good in thinking any trifles that concern [him] are so much worth of [his] attention’ and gives him ‘entirely liberty to make what Additions [he] please[s] to the account of [his] Life’ (LDH 2.336, 540). Strahan tells Smith that the Addition will be ‘highly proper’, like ‘every particular respecting that great and good man’.50 At the beginning of October Smith’s Addition is ready: ‘I think there is a propriety in addressing it as a letter to Mr Strahan to whom [Hume] has left the care of his works.’51 Acknowledging that Smith’s narrative is the consequence of his ‘request’ and Hume’s ‘approbation’ of it, Hume’s brother ‘much’ approves it and suggests a few alterations, consistent with the ‘short and simple a manner’ of the Life.52 Smith adopts his remarks and sends Strahan his ‘small addition’ to Hume’s ‘small piece’ (LDH 2.318, 522A; 2.323, 525), and Strahan likes it ‘exceedingly’.53 In 1777, contrary to Hume’s dispositions, Strahan publishes the Life and the Addition ‘separately’ from Hume’s work. The Dialogues are published in 1779 (three years after his death) by Hume’s nephew, possibly with some help from Strahan. According to friends and doctors, in his last months Hume ‘amuses himself’ with reading (classic) ‘amuseing books’.54 And in his last days he is reading ‘the dialogues of Lucian’, the ‘Dialogues of the Dead’ and ‘the dialogue entitled Kataplu’. On 8 August 1776 Hume has a Lucianic conversation with himself. By mixing up Megaphentes’ excuses (‘Cataplus’) with Socrates’ final resistance (Dialogues of the Dead), he invents ‘several jocular excuses’

to obtain a delay from Charon.55 As a writer very busy in ‘correcting [his] works for a new edition’, Hume asks for ‘a little time’ to see ‘how the public receives the alterations’; as an anti-religious writer ‘endeavouring to open the eyes of people’ he asks for ‘a few years longer’ to have ‘the satisfaction of seeing the downfall of some of the prevailing systems of superstition’. Charon is not convinced, Hume wants ‘a lease for so long a term’: ‘there will be no end’ of correcting and opening people’s eyes will take ‘many hundred years’.56 By the way, Lucian – says the Enquiry – ‘entirely opened the eyes of mankind’ (EHU 10.23 / 121). Hume, says Smith, ‘diverts himself with correcting his own works’, he makes ‘many very proper corrections, chiefly in what concerns the language’, and Smith is ready to ‘revise the sheets and Authenticate its being according to his last corrections’, as he has ‘promised’ to Hume.57 In the Addition Smith prepares the excuse of correcting by telling the reader that Hume ‘continue[s] to divert himself, as usual, with correcting his own works for a new edition’.58 This agrees with Hume’s self-representation as an author extremely ‘anxious of Correctness’ (LDH 1.175, 82; 2.304, 511; see ibid. 1.38, 16; 1.175, 82; 2.239, 455). ‘There is no End of correcting’, proclaims Hume in 1763, and in 1771, ‘an Author may correct his works, as long as he lives’ (LDH 1.379, 203; 2.246–7, 457). As he tells the printer, ‘I am perhaps the only Author . . . who gratutiously employ’d great Industry in correcting a Work, of which he has fully alienated the Property’ (LDH 2.239, 455). Hume, who finds ‘curious’ that an author could have no patience to read over his published works (NLH 62, 31; LDH 2.31, 314), recalls in the 1770s a ‘saying’ of Jean-Baptiste Rousseau the poet and offers his own version: 31

9780826443595_Ch01_Final_txt_print.indd 31

2/9/2012 9:39:44 PM

HUME’S LIFE any further Improvements. This is some small Satisfaction to me in my present Situation; . . . it is almost the only one that my Writings ever afforded me: For as to any suitable Returns of Approbation from the Public . . . they are yet to come (LDH 2.322, 525).

‘a man might spend his whole Life in correcting one small Volume, and yet have inaccuracies in it’ (LDH 2.243, 456; 2.304, 511; 2.250, 461).59 In 1742 he makes a mistake in quoting the aforementioned Rousseau (‘C’est la politesse d’un Suisse / Dans la Hollande civilisé’), and six years later he readily corrects it (‘En Hollande civilisé’).60 Most of his corrections ‘fall upon the Style’, he easily allows, and locating and eliminating inaccuracies gives him a ‘sensible pleasure’ (LDH 2.151, 394; 2.188, 422; 2.243, 456; 2.250, 461). At the end of July 1776 he employs himself in correcting the sixth volume of the History and the moral Enquiry. It is the trifling occupation of a dying author:

In his Lucianic dialogue Hume is concerned with the effect of his writings on religion. In his Addition Smith tries to soften the effect of this imaginary conversation with the mythical ferryman: he moderates the language and introduces the excuse that Hume desires yet a further opportunity to correct his manuscripts. According to Smith’s letter to Wedderburn, Hume is dying with ‘great chearfulness . . . and more real resignation to the necessary course of things, than any Whining Christian ever dyed with pretended resignation to the will of God’; according to the Addition, he is simply dying with ‘the utmost cheerfulness, and the most perfect complacency and resignation’.62 Hume’s ‘pleasure of seeing the churches shut up, and the Clergy sent about their business’ is turned into that of ‘seeing the downfall of some of the prevailing systems of superstition’.63 Hume’s brother remarks that Hume did not say ‘I am dying as fast as my worst enemies could wish’, as Smith writes in the first draft of the Addition, but ‘as my enemies, if I have any, could wish’.64 Smith accepts the correction and adds that Hume has ‘no enemies upon whom he wished to revenge himself’.65 This fits with My Own Life, where Hume (ironically and by a marginal addition) declares that he is ‘little susceptible of enmity’ and has never been attacked by the ‘baleful tooth’ of ‘calumny’, even though he ‘wantonly’ exposed himself to ‘the rage of both civil and religious factions’ (MOL, LDH 1.7). The Life sounds

You will wonder, that, in my present Situation I employ myself about such Trifles, and you may compare me to the modern Greeks, who, while Constantinople was besieged by the Turks and they themselves were threatened with total Destruction, occupyed themselves entirely in Disputes concerning the Procession of the holy Ghost. Such is the Effect of long Habit! (LDH 2.329, 531) Strahan takes these lines as a ‘living Evidence’ that we are ‘much interested in what is to pass after our Deaths’.61 In his last letter to Strahan, concerning the moral Enquiry, Hume announces: ‘This is the last Correction I shall probably trouble you with: . . . all shall be over with me in a very little time.’ (LDH 2.331–32, 537) Hume’s excuses in Smith’s Addition sound like a little pleasantry about his own final remarks: I have made [my new edition] extremely correct; . . . if I were to live twenty Years longer, I shoud never be able to give it 32

9780826443595_Ch01_Final_txt_print.indd 32

2/9/2012 9:39:45 PM

HUME’S LIFE Volpone, talk of childless rich old men and parasitical clients rather than real friends: perfect friendship does not contemplate money, as Lucian remembers in ‘Toxaris’ and Montaigne in ‘L’Amitié’. In Hume’s ‘A Dialogue’, ‘by his Will’ Usbek makes his ‘intimate Friend’ Alcheic ‘his Heir to a considerable Part of his Fortune’ (EPM Dial. 8 / 326). By his own will Hume leaves 200 pounds sterling to his ‘friends’ Ferguson, d’Alembert and Smith. D’Alembert accepts the legacy. He had helped Hume with the publication of the original French version of A Concise and Genuine Account of the Dispute between Mr. Hume and Mr. Rousseau, and Hume had even asked Smith to tell d’Alembert he had made him ‘absolute Master to retrench or alter what he thinks proper’ (LDH 2.83, 348).69 Smith did not help Hume with the correction and publication of the Dialogues: he discharges the legacy because he cannot ‘with honour accept it’.70 Hume had ‘hitherto forborne’ to publish the Dialogues, desiring to ‘keep remote from all Clamour’ (LDH 2.323, 525); Smith is ‘uneasy about the clamour’ they could excite. Worried that they could ruin his tranquillity, he tries to make Strahan apprehensive that they could ruin his interest.71 Hume assures Smith that the Life is a ‘very inoffensive’ piece (when compared to the Dialogues), which ‘will be thought curious and entertaining’ (LDH 2.318, 522A; 2.323, 525); Smith agrees, and is later offended by the reception accorded his own Addition: ‘a single, and as I thought a very harmless Sheet of paper, . . . concerning the death of our late friend Mr Hume, brought upon me ten times more abuse than the very violent attack I had made upon the whole commercial system of Great Britain’.72 In 1747 Hume says he could not see ‘what bad consequences follow, in the present

like the Manifesto of the Humean Party: ‘English, Scotch, and Irish, Whig and Tory, churchman and sectary, freethinker and religionist, patriot and courtier, united in their rage against [me]’ (MOL, LDH 1.4; also see NLH 194, 107). Before publishing the third Book of the Treatise, Hume had already realized that the clergy are ‘always Enemys to Innovations in Philosophy’, and only after the publication of the History does he rather belatedly resolve ‘to be more cautious than formerly in creating myself Enemies’ (LDH 1.37, 15; 1.352, 189).66 As Strahan realizes and Smith endeavours to forget, Hume shows an ‘extreme solicitude’ to publish the Dialogues.67 He has been accused of being ‘as great an atheist as Bolingbroke’, even though Bolingbroke’s volumes contain ‘so little Variety & Instruction’ and the clergy have ‘no Reason’ to be ‘enrag’d against him’ (LDH 1.168, 78; 1.209, 101; 1.214, 105). Yet, for the sake of his Dialogues, Hume does not scruple to compare himself to Bolingbroke. And he tells both Smith and Strahan that Bolingbroke’s editor was not ‘any wise hurt by his Publication’, as ‘he always justify’d himself by his sacred Regard to the Will of a dead Friend’. If he leaves Strahan the Dialogues ‘by Will’, his ‘executing the Desire of a dead Friend, will render the publication still more excusable’ (LDH 2.316, 522; 2.324, 525). According to the last codicil, the ‘duty’ of his nephew in publishing the Dialogues ‘as the last Request of his Uncle, must be approved of by all the World’.68 In the 1750s Hume is extracting from the classics ‘what serv’d’ concerning the ancients (LDH 1.152–3, 71): from Lucian, he says, we may ‘gather’ that ‘leaving great sums of money to friends’ was a common practice in Greece and Rome (E 400n), even though the Dialogues of the Dead, like Jonson’s 33

9780826443595_Ch01_Final_txt_print.indd 33

2/9/2012 9:39:45 PM

HUME’S LIFE says Strahan, ‘knew [Hume] well, and loved him much’, and it ultimately falls to Smith to tell Strahan and the world of the death of ‘our most excellent, and never to be forgotten friend’. After 11 years he also unforgivably celebrates ‘the abilities and virtues of the never to be forgotten Dr Hutcheson’.76

age, from the character of an infidel’ (LDH 1.106, 58); 30 years afterwards Horne vituperatively attacks the ‘foolish and insensible’ Hume for reading Lucian and inventing droll conversations with Charon. Hume’s excuses are an attack on ‘superstition’, and ‘we all know . . . against what Religion his shafts are levelled, under that name’. Smith wished that the Addition could be agreeable; Horne thinks that he who can reflect ‘with complacency’ on Hume, amusing himself at his death ‘may smile over Babylon in ruins’.73 Hume was reading Lucian. ‘It is an idle thing in us,’ he writes in June, 1776, ‘to be concerned about any thing that shall happen after our Death’; yet, he adds, ‘I often regretted that a Piece, for which I had a particular Partiality, should run any hazard of being suppressed after my Decease’ (LDH 2.325–6, 527). Again Strahan takes it as a sign that ‘our Existence will be protracted beyond this life’.74 With regard to the Enquiry, in 1754 Hume had allowed: ‘I am willing to be instructed by the Public; tho’ human Life is so short that I despair of ever seeing the Decision’; in 1776 he laments: ‘I shall not live to see any Justice done to me’ (LDH 1.187, 91; 2.322, 525). ‘If I live a few Years longer,’ Hume tells Smith, ‘I shall publish [the Dialogues] myself’ (LDH 2.316, 522); and Smith tells Strahan: ‘I could have wished [they] had remained in Manuscript to be communicated only to a few people . . . [they] never should have been published in my lifetime.’ In 1763 Elliot and Blair had expressed the same sentiments.75 In his last days Hume was ‘revising’ the Dialogues (LDH 2. 334, 538), but despite these exertions on the part of a dying man his good friend Smith remained determined to have nothing to do with the publication of what he saw as so incendiary a piece of writing. Smith,

9. STUCK IN A BOG (FISHWOMEN FOR THEOLOGIANS) There is an anecdote about one of Hume’s misadventures in Edinburgh, which much amused the historian Sir Leslie Stephen and his daughter Virginia Woolf, and reminds us of Thales’ tumble and the laughter of the Thracian maid.77 It agrees with the pagan frolic tone of the Natural History and says something about Hume’s style, his attitude towards religion, and his not-so-serious pronouncements on women, monks and superstition (NHR 12.144). It also says something about the distinction between popular and pretended philosophical religion, and some Humean resolutions to ‘keep [him]self in a proper disposition for saying the Lord’s Prayer, whenever [he] shall find space enough for it’ and to ‘proceed directly to attack the Lord’s Prayer & the ten Commandments & the single Cat’ (LDH 1.148, 70; NLH 43, 25). Finally, it says something about his real expectations as regards the effect of his work: ‘this did not convert the generality of mankind from so absurd a faith; for when will the people be reasonable?’ (NHR 12.173). Hume is engaged in building a ‘small House, I mean [he says] a large House for an Author’ (he did finish it), which is the ‘second great Operation of human Life’ (LDH 2.232, 451; 2.235, 453). As always, he is out of the common road: 34

9780826443595_Ch01_Final_txt_print.indd 34

2/9/2012 9:39:45 PM

HUME’S LIFE On his daily visits to inspect the work, he was in the habit of taking a short cut across what was then a swamp . . . he made a slip, fell over, and stuck fast in the bog. Observing some Newhaven fishwomen passing with their ‘creels’, he called aloud to them for help; but, when they came up, and recognised the wicked unbeliever David Hume, they refused any assistance, unless he first repeated, in a solemn tone, the Lord’s Prayer. This he did, without pause or blunder, and was extricated accordingly. He used to tell this story himself with great glee, declaring that the Edinburgh fishwives were the most acute theologians he had ever encountered.78

4

5

6 7

8

Even when getting out of his bog, the sceptic will always be ‘the first to join in the laugh against himself’ (EHU 12.23 / 160).

9

10

NOTES

11 12

1

2

3

Early Responses to Hume’s Life and Reputation, 2 vols, 2nd edn, ed. J. Fieser (Bristol: Thoemmes Continuum, 2005), vol. 1, p. 294. These alleged words are often treated as an indication that Hume’s mother viewed him as weak-minded; other interpreters maintain that the sense of the words in the local ‘Vernacular’ of the time was such that his mother was drawing a contrast between his good nature and his ‘uncommon acuteness’ (see H. Calderwood, David Hume (Edinburgh: Oliphant Anderson & Ferrier, 1898), p. 14; D.F. Norton, ‘An Introduction to Hume’s Thought’, in D.F. Norton (ed.), The Cambridge Companion to Hume (Cambridge: Cambridge University Press, 1993), pp. 1–32, p. 2; E. Mazza, ‘La mamma di Hume. Interpretazioni di un detto apocrifo’, in Il mestiere di studiare e insegnare filosofia (Milano: Wise, 2000), pp. 93–152). J.H. Burton, Life and Correspondence of David Hume, 2 vols (Edinburgh: W. Tait,

13

14

15

16 17

18

19 20 21

1846), vol. 1, p. 294n1; S.J. Pratt, Supplement to the Life of David Hume (London: J. Bew, 1777), pp. 33–4. Hume seems to reflect on La Rochefoucauld’s maxim: ‘everybody speaks well of his own heart, nobody dares to speak well of his own understanding’ (La Rochefoucauld, Maximes (Paris: Garnier, 1991), p. 309. Letters of Eminent Persons addressed to David Hume, ed. J.H. Burton (Edinburgh: W. Blackwood, 1849), p. 252. Burton, Life, vol. 1, p. 226. J. Brown, An Estimate of the Manners and Principles of the Times, 2nd edn (London: L. David and C. Reymers, 1757), p. 57. E.C. Mossner, ‘New Hume Letters to Lord Elibank, 1748–1776’, Texas Studies in Literature and Language 4 (1962), p. 445; A Concise and Genuine Account of the Dispute between Mr. Hume and Mr. Rousseau (London: T. Becket and P.A. De Hondt, 1766), pp. iii, iv, viii. Horace Walpole’s Correspondence with Madame Du Deffand and Wiart, ed. W.S. Lewis and W.H. Smith (New Haven: Yale University Press, 1970), pp. 253, 266. BL, Add. MS 36638, f. 22r. NLS, MS 25692, f. 33v. E.C. Mossner, ‘Hume at La Flèche, 1735: An Unpublished Letter’, University of Texas Studies in English 37 (1958), pp. 30–3, p. 32. W. Leechman, The Temper, Character, and Duty of a Minister of the Gospel, 5th edn (Glasgow: R. Foulis, 1749), pp. 34, 41. J.C. Hilson, ‘More Unpublished Letters of David Hume’, Forum for Modern Language Studies 6(4) (1970), pp. 315–26, p. 321. LDH 1.131, 64; NLS, MS 25708, f. 36v; MS 25703, f. 187v. NLS, MS 25703, ff. 188r, 212v. B. Mandeville, The Fable of the Bees, ed. P. Harth (London: Penguin, 1989), p. 341. NLS, MS 25703, 188r; ff. 210r, 212v; Early Responses to Hume’s Life and Reputation, ed. J. Fieser, vol. 2, p. 203; Charlemont, ‘Anecdotes of Hume’, RIA, Charlemont MSS 12.R.7, f. 518v. NLS, MS 25708, ff. 40v; LDH 1.132, 64. NLS, MS 25708, ff. 46rv, 50v. NLS, MS 25708, ff. 33r, 41v; E 2.7.15, 339;

35

9780826443595_Ch01_Final_txt_print.indd 35

2/9/2012 9:39:45 PM

HUME’S LIFE

22 23 24

25 26 27 28 29

30

31 32

33

34

35

36

37

38 39

40

41

Home, J., ‘Diary of a journey with Hume to Bath’, in Early Responses, vol. 1, p. 284; NLS, MS 25708, f. 33r. Letters of Eminent Persons, p. 165. Ibid., p. 67. D. Diderot, Correspondance V, ed. G. Roth, trans. E. Mazza (Paris: Les Editions de Minuit, 1959), p. 132. NLS, MS 23163, f. 100. NLS, Acc 11139, f. 24r. NLS, Acc 11139, f. 25r. A Concise and Genuine Account, p. 90. Charlemont, ‘Anecdotes of Hume’, f. 521; Horace Walpole’s Correspondence with Hannah More et al., ed. W.S. Lewis et al. (New Haven: Yale University Press, 1961), pp. 47, 49. Mossner, ‘New Hume Letters to Lord Elibank’, p. 452. NLS, MS 23159, item 23, f. 10. L. Sterne, Letters, ed. L.P. Curtis (Oxford: Clarendon, 1967), pp. 157, 219. Sterne, Letters, p. 218; L. Sterne, ‘The Case of Hezekiah and the Messengers’, The Sermons of Mr. Yorick, vol. 3 (London: T. Becket and P.A. De Hondt, 1766), pp. 30, 42. Sterne, Letters, pp. 219, 239, 445 (also see pp. ivi, 235, 254); L. Sterne, A Sentimental Journey through France and Italy by Mr. Yorick (London: T. Becket and P.A. De Hondt, 1768), pp. 92–3. Lettres d’André Morellet. Tome I: 1759–1785, ed. D. Medlin et al. (Oxford: The Voltaire Foundation, 1991), p. 72. The Correspondence of Dr William Hunter, 1740–1783, ed. C.H. Brock, 2 vols (London: Pickering & Chatto, 2008), vol. 1, pp. 229, 230–1. Letters of Eminent Persons, pp. 171–72; see J. Curry, An Historical Review of the Civil Wars in Ireland, 2 vols (Dublin: L. White, 1786), vol. 1, p. 215n. NLS, MS 23156, f. 77. Dublin, Trinity College, 543/2/11 /E.2.19; E. Murphy, The Select Dialogues of Lucian (London and Dublin: Edward Exshaw, 1746), pp. xi–xii. Notes and Queries 3 (8 April 1899), p. 262a. T. Kozanecki, ‘Dawida Hume’A Nieznane Listy W Zbiorach Mezeum Czartoryskich (Polska)’,

42

43

44 45

46 47

48 49

50 51 52 53 54

55

56 57 58 59

60

61 62

63 64

65

Archiwum Historii Filozofii I Mysli Spolecznej 9:5 (1963), p. 138; LDH 1.10, 1. The Correspondence of Adam Smith, ed. E.C. Mossner and I.S. Ross (Indianapolis: Liberty, 1987), p. 168. D. Hume, ‘Disposition and Settlement’, 4 January 1776, in ‘Record of Testaments for the year 1781’ (8 March 1781), NAS, CC8/8/125, ff. 863, 865. The Correspondence of Adam Smith, p. 211. D. Hume, ‘Codicil to my Will’ (7 August 1776) (1st codicil: ‘In my latter Will and Disposition’; 2nd codicil: ‘I also ordain’), NLS, MS 23159, item 24, f. 17; see LDH 2.318, 522A; 2.323, 525. NLS, MS 23159, item 24, f. 17; LDH 2.323, 525. The Correspondence of Adam Smith, pp. 199, 206, 211. Ibid., pp. 203–4, 206. Hume, ‘Disposition and Settlement’, f. 863; The Correspondence of Adam Smith, p. 206; see ibid., pp. 203, 219. The Correspondence of Adam Smith, p. 212. Ibid., pp. 214, 216. Ibid., pp. 206, 208, 214–15. Ibid., pp. 216, 221, 222. Ibid., pp. 207, 218; J. Home, ‘Diary of a Journey with Hume to Bath’, p. 281. The Correspondence of Adam Smith, pp. 203, 219; Correspondence of Dr William Hunter, vol. 2, p. 226. The Correspondence of Adam Smith, p. 219. Ibid., pp. 204, 206, 211. Ibid., p. 218. J.-B. Rousseau, ‘Preface’ (1712) to Œuvres diverses, 2 vols (London: J. Tonson and J. Watts, 1723), vol. 1, p. iv. E 127. The initial error and subsequent correction are displayed in D. Hume, Essays, Moral and Political (Edinburgh: A. Kincaid, 1742), p. 80, and Essays, Moral and Political, 3rd edn, corrected (London: A. Millar; Edinburgh: A. Kincaid, 1748), p. 177. NLS, MS 23157, item 69, f. 294. The Correspondence of Adam Smith, pp. 203, 218 (see also p. 206). Ibid., pp. 204, 219. Ibid., p. 215 (see Correspondence of Dr William Hunter, vol. 2, p. 226). The Correspondence of Adam Smith, pp. 215, 219.

36

9780826443595_Ch01_Final_txt_print.indd 36

2/9/2012 9:39:45 PM

HUME’S LIFE 66 67 68 69

70

71 72 73

74

75

See Burton, Life, vol. 1, p. 226. The Correspondence of Adam Smith, p. 212. NLS, MS 2319, item 24, f. 17. Letters of Eminent Persons, p. 252; J.C. Hilson and J.V. Price, ‘Hume and Friends, 1756 and 1766: Two New Letters’, The Yearbook of English Studies 7 (1977), pp. 121–7, p. 126. The Correspondence of Adam Smith, pp. 209, 214, 210, 215. Ibid., pp. 211, 216–17. Ibid., p. 251. G. Horne, A Letter to Adam Smith LL.D. On the Life, Death and Philosophy of his Friend David Hume (Oxford: Clarendon Press, 1777), pp. 9–11, 12–13, 29–30. NLS, MS 23157, item 69, f. 297.

76 77

78

The Correspondence of Adam Smith, p. 211; LDH 1.380, 203; NLH 71, 35; 72, 36; NLS MS 23153, 51, f. 157. The Correspondence of Adam Smith, pp. 220, 309. L. Stephen, ‘Hume’, in S. Lee (ed.), Dictionary of National Biography (London: Smith, Elder, 1891), vol. 28, p. 223b; V. Woolf, To the Lighthouse (1927) (London: Grafton, 1988), pp. 62, 70; G.L. Strachey, ‘Hume’ (1928), in Portraits in Miniature and Other Essays (London: Chatto & Windus, 1931), pp. 151–52. Selections from the Family Papers Preserved at Caldwell, pt II, vol. 2 (Glasgow: 1854), pp. 177–8n1; see The Scotch Haggis (Edinburgh: D. Webster and Son, 1822), p. 77; Burton, Life (1846), vol. 2, p. 458.

37

9780826443595_Ch01_Final_txt_print.indd 37

2/9/2012 9:39:46 PM

2 HUME’S EMPIRICIST INNER EPISTEMOLOGY: A REASSESSMENT OF THE COPY PRINCIPLE Tom Seppalainen and Angela Coventry

1. INTRODUCTION

2. COPYING AND ITS TARGETS

Vivacity, the ‘liveliness’ of perceptions, is central to Hume’s epistemology. Hume equated belief with vivid ideas. Vivacity is a conscious quality and so believable ideas are felt to be lively. Hume’s empiricism revolves around a phenomenological, inner epistemology.1 Through copying, Hume bases vivacity in impressions. Sensory vivacity also concerns liveliness or patterns of change. Through learnt skilful use, vivacity tracks change specific to intentional sense-perceptual experience consisting in Hume’s ‘coherent and constant’ complex impressions. Copying, in turn, communicates vivacity to ideas where it becomes an indicator of the believability of ideas. Hume’s copying concerns then the causation of conscious skills required for the identification of empirically warranted structures. Copying allows Hume to combine a radically externalist empiricism with a phenomenological inner epistemology.

The copy principle connects ideas to impressions. The actual process is left without much clarification by Hume. We know it is a causal one and that impressions are causally prior to ideas. We also know that causation holds only for concrete particulars. These ideas, together with Hume’s general scepticism about the intelligibility of causal relations, suggest that a lack of insight into the process is to be expected. From Hume’s standpoint, there are not many empirically founded conceptual insights to share. Although such a state of affairs is common for fundamental posits of scientific theories including those of psychology, it has caused much dismay in philosophy and for a good reason. At strategic philosophical occasions, when Hume challenges the intelligibility of commonly accepted philosophical beliefs, he introduces copying as a test or justificatory principle (see, for example, EHU 1.13 / 13). Copying is Hume’s distinct operationalization of

38

9780826443595_Ch02_Final_txt_print.indd 38

2/9/2012 9:39:33 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY Once we articulate the targets of copying, we can form an exact judgement about Hume’s externalist empiricism. Prima facie, copying indicates externalism. Copies share significant properties with their originals. Think of scale models and photocopies. Both are results of copying and both appear like their originals. The terminology Hume uses indicates a commitment to this strong externalist form of copying. Ideas for him are ‘faint’ versions of impressions (THN 1.1.1.1 / 1; EHU 2.5 / 19). Copies can also function like their originals and not just seem or appear like them. Often they do both. This point echoes through the history of debates between internalists and externalists. One philosopher’s empiricism is another’s nativism when it is not specified whether properties or processes or both are shared across systems. The historical reminder is important in the case of Hume. He uses potentially misleading terminology according to which ‘perceptions of the mind’ are ‘objects.’ But copying need not be limited to the passing of objects and their properties to cognitions for processing. All kinds of ‘entities’ can be copied, including functions or activity types. We pose the question about the targets of copying in an open-minded way. According to collective wisdom, copying indicates Hume’s ‘meaning empiricism’. According to this ‘semantic hermeneutics’, linguistic meaning is the target of copying. The interpretation is codified in slogans: copying gives the ‘empirical cash value’ of terms.3 Hume’s concern with the meaning of linguistic expressions has textual support. He claims to apply the copy principle to philosophical terms (EHU 2.9 / 22). Despite this, the semantic hermeneutics distorts the causal nature of copying. Copying becomes

empiricism; the paucity of information on the process leaves Hume’s empiricism with a ‘conceptual’ problem. This is the context for our reassessment of the copy principle. Despite the conceptual problem, copying is the centerpiece of Hume’s empiricism. Why this is so can be best explained in programmatic terms. The distinguishing features of the few broad epistemological orientations are clear.2 Empiricism embodies an externalist explanatory strategy for whatever it seeks to explain or clarify. In Hume’s case, the copy relationship establishes an externalist intellectual strategy for the cognitive realm of ideas through the sensory realm of impressions. Externalism differs from internalist paradigms of epistemology such as rationalism and constructivism. These emphasize the role of inherent, innate, or just a priori posits for cognitive and behavioural outcomes. Empiricism minimizes the significance of such posits even if the mind is not construed as a Lockean tabula rasa. Epistemological internalism is not the only relevant contrast. Externalism comes also in a form that couples the mind (or parts of it) to the world or environment (or parts of it). This type should be called ‘objectivism’. Hume’s ‘externalist’ empiricist epistemology concerns the role of the sensory environment for acts of cognition. Copying is a relation between two mental domains and does not entail objectivism. In Hume’s case, objectivism follows only if the sensory system is explained by factors of the environment, ones external to it. We leave open the issue of objectivism. An externalist hypothesis about Hume’s empiricism leads to a straightforward heuristic for clarifying copying. Since the process is unclear, we must focus on its targets.

39

9780826443595_Ch02_Final_txt_print.indd 39

2/9/2012 9:39:33 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY are copied, senses and cognitions resemble in terms of their structural features. This holds even if no ‘abstract’ structure ever occurs in the systems independently, without ‘objects’ or contentful complex information. Structure need not be independent for it to be a psychologically and epistemologically critical factor. It only needs to occur robustly across circumstances, have a cognitive function and be knowable. The robust presence of particular complex impressions is enough for the first condition. The latter two conditions will be taken up next. Copied structure opens up the possibility for further targets of copying. Structure itself has an origin. This may be similar for impressions and ideas, at least under some circumstances. Generally, if complex structure is the end result – effect – of some sort of activity – its cause – such activity may itself be copied. More specifically, if the structured information in the senses is a function of activity, Hume’s system allows for its copying. Since Hume’s psychology is a ‘genetic’ or developmental theory this is a genuine possibility. We believe that it is more than a theoretical possibility. The fact that Hume never discussed the copying of activities should not prejudice this theoretical option. A direct treatment is prohibited by Hume’s scepticism about the intelligibility of all causal relations qua relations. It applies to activity. For this reason, Hume included in his psychological taxonomy an inherently action-related quality, ‘vivacity’, the ‘liveliness’ of perceptions. Hume’s synonyms, ‘strength’, ‘violence’ and ‘vigour’, are all action-related terms. To explore the copying of activity we must address the copying of vivacity. Hume never clearly mentions the copying of vivacity. It is obvious in the light of his methodology and theory. Targets of copying

interpreted in analytic or definitional terms such as operational definitions. This is not Hume but empiricism of a much later origin. And semantic hermeneutics is not just bad historiography. It has led to scathing critiques of Hume. Although we believe in history serving current concerns, these critiques should by now be seen as vitiating any such edifying function for the semantic hermeneutics. To us, they raise only one question. Why study Hume if his central causal principle is without merit, theoretical function, or promise? For us, copying targets information of the senses and information comes in many forms, including semantic ones. Let us start with Hume’s ideas about the targets of copying that fit clearly under the heading of information and then turn to the less obvious informational targets. Hume’s senses include atomic elements or ‘simple impressions’. Any particular impressions of sight, smell, taste, sound, touch, pleasure or pain, are copied to the realm of ideas as simple ideas. But this is not all. Most information is in a form more complex or structured than atomic perceptions and this is also the case for Hume. Both senses and cognitions contain complex impressions and ideas, respectively. Some of this structured information of cognition owes its origin to copying but not all of it. Hume does not say whether all structured information in the senses is copied to cognitions; it is clear that ‘cognitive’ processes of imagination generate complex information that is novel relative to the senses. It is novel since no such specific complex structures exist in the senses – even if the elemental information does. The idea of a unicorn is the classic exemplar. So far everything is clear and uncontroversial. The copying of even some complex impressions shows that copying also targets structure. Given that some complex impressions 40

9780826443595_Ch02_Final_txt_print.indd 40

2/9/2012 9:39:33 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY role. Hume uses relative vivacity in psychological taxonomy to classify perceptions into ideas and impressions, distinguish memory from other ideas, and distinguish beliefs from mere fictions of the imagination. Thus, through vivacity, information about mental activity acquires a fundamental taxonomic role in Hume’s psychology. Even if we cannot know activity itself, we can know its quality, vivacity. Vivacity is the knowable indicator of activity. But this is not all. Hume equated the central epistemological notion, belief, an idea type, with vivacity (THN 1.3.9.15 / 106). This has caused much puzzlement in the secondary literature.4 Much of it can be resolved by articulating the informational nature of vivacity. ‘Belief’ is an ambiguous term and combines information with attitude. When we believe some X we ‘have’ the informational content X ‘in our mind’. But we also ‘have’ a specific doxastic attitude towards the content (THN 1.3.7.1 / 94). ‘Belief’ is normally used for both. In our analysis, ‘belief’ refers to the former and ‘believability’ to the latter, doxastic attitude. For Hume, vivacity is a quality of perceptual objects instead of the object as such. Thus, vivacity should not be identified with belief as such. Instead, vivacity concerns believability. Believability is based in the ‘lively’ or active dimension of ideas. Since vivacity is copied, the doxastic attitude of believability is copied. The copying of both vivacity, a quality of perceptual objects, and the informational content of perceptual objects couples belief-ideas tightly with the senses. This is understandable for an empiricist. For rationalists, ‘judgement’ stands behind ‘believability’ as an independent internal epistemological process. In Hume’s externalism, both the doxastic attitude and (much of) the contents of belief are copied to cognitions from the realm of the senses.

can be explored through a resemblancebased interpretative methodology. This coheres fully with Hume’s theory and use. For Hume, ideas are not only caused by but also resemble impressions (THN 1.1.1.6 / 4). Copying results in a resemblance between impressions and ideas as ‘objects’. Moreover, if ideas are particular images, then ideas are literal ontological copies. Copying can also concern activities and still be operationalized in terms of resemblance. And finally, copying is an associative causal process yet unique in being based in the ‘natural’ relation of resemblance instead of contiguity. Importantly, a resemblance methodology can be used in a two-way fashion. Facts concerning the realm of impressions can be used to interpret the realm of ideas. Nothing prohibits using resemblance in reverse, in theorizing and interpretation. Hume arguably did both if only because ordinary consciousness is ambiguous between the presence and contribution of ideas and impressions. Let us apply this methodology to vivacity. Impressions and ideas resemble in terms of vivacity in the way that ideas and impressions generally resemble. Idea copies are ‘faded’ versions of impressions also with respect to the quality of vivacity. Clearly, vivacity is a target of copying. But is it not already included in the general idea of a faded objectcopy? The question points at the information carried by vivacity. That is of a kind no individual ‘perceptual object’ can contain. Let us see why. For Hume, vivacity is a ‘quality’ of perceptions. This does not mean that vivacity is an ordinary ‘inherent’ property. It cannot be transferred to the realm of ideas with a copied object. As liveliness, ‘vivacity’ is a quality of a ‘series’ of impressions and of the changes between them. Slow change is not lively. As a quality of change, liveliness has an indicator 41

9780826443595_Ch02_Final_txt_print.indd 41

2/9/2012 9:39:33 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY Descartes’s methodology of clear and distinct ideas is the famous ‘inner’ or phenomenological epistemology. Despite the ‘Rules for the Direction of the Mind’,5 the conscious skills of thinking needed to attain clear and distinct ideas remain unsatisfactory. The flavour of a pseudo-methodology, a mere ‘inner gaze’ without a clear learning history, remains. The reason is in the innate nature of Cartesian epistemologically appropriate ideas. Descartes’s metaphysics of consciousness complements the epistemology. Consciousness is independently in all mental systems, governing and unifying them. Hume’s metaphysics is externalist – ‘genetic’ or developmental and his ‘inner’ epistemology complements it. His emphasis on ‘experience’ and external causal factors opens the possibility that epistemologically relevant conscious skills are learnt through a specific causal history. Vivacity has a central role in the specific causal history of skills of conscious activity. Tracking change in perceptions through vivacity is of epistemological use only if vivacity discriminates among different types of change. More specifically, vivacity allows for distinguishing beliefs from mere conceptions only if we are able to ‘read’ change. How should this skill be conceptualized? Copying makes it possible that we learn from vivacity within the sensory realm and then use it within the realm of cognition. So what does sensory vivacity teach us? Sensory vivacity also tracks change. One ‘type’ of sensory change has a clear cognitive function for Hume. Complex impressions that have the ‘qualities’ of coherence and constancy are involved in the causation of the belief in an external world (THN 1.4.2.15–20 / 194–5). And this belief-idea, in turn, is conscious information about the existence of a world independent of us. Thus, only some complex impressions through copying

Above we dubbed vivacity an ‘indicator’ for psychological taxonomy. The idea applies also to belief. The quality of vivacity is an indicator of believability. Vivid ideas cause the doxastic attitude. This shows that structured information has a central cognitive role for Hume. Through its origins, that role extends to the activities responsible for structure. The second condition for taking complex structure as a target of copying is satisfied. To complete the sketch, we must address knowledge of structure and its origins. Vivacity is a feeling for Hume. This is not surprising. All central posits of his mental mechanism have a conscious dimension. Yet vivacity is unique among these. It is a qualitative indicator of activity. Hume’s scepticism about ‘relations themselves’ covers all activities. It also spans all methods including ‘inner’ ones. The evidentiary limits of phenomenology for activities are severe and comprehensive. No wonder Hume aspired to be the Newton of the Mind (THN Intro. 7–8 / xvi–xvii). He professed no comprehensive phenomenological epistemology for the ‘metaphysical’ nature of mental associative relations that make the mind move; with his ‘non fingo’, Newton professed the same for all observational methodologies concerning force or gravitational ‘associations’ that make objects with mass move. Vivacity remains ‘only’ an indicator of activity. There is a further limit. All perceptual objects are vivid, yet, for Hume, not all of them are believable. Vivacity is only ‘effects-based’ information about activities, yet sufficient to discriminate among critical activity types. Vivacity is a conscious indicator whose use or ‘indicator function’ must be learned. It is a skill of conscious activity or conscious skill, for short. The idea of a learnt conscious epistemic skill is best introduced through its alternative, Cartesianism. 42

9780826443595_Ch02_Final_txt_print.indd 42

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY elements through structure to activities and conscious qualities that track them constitutes the targets of Humean copying. As a result, copying grounds Hume’s epistemology causally and creates an externally governed inner phenomenological epistemology. To bolster the idea that copying communicates to ideas a conscious indicator of believability, for the remainder of our chapter we focus on the secondary literature. We analyse four interpretations of vivacity directly relevant to our sense-based account. Each has a kernel of truth yet each fails in an instructive way. The critical analysis is intended both to clarify and support our position. If it fails to do so, a critical review of the secondary literature stands on its own and fits well with the spirit of a reassessment.

partake in the causation of the intentionality or ‘about-ness’ of ideas. Through their phenomenological nature and this causal role, coherence and constancy themselves are conscious sensory information about the world. The skilful use of vivacity is learnt in this context. The correlation between different senseperceptual outcomes, coherent and constant complex impressions and their opposites, and different profiles of sensory vivacity grounds vivacity’s indicator function in the senses. The conscious skill is learned in the senses. More specifically, in the copied form that indicates believability, vivacity informs about those changes in ideas that are relevant for empirically warranted believability. In the senses, such changes cause sense-perceptual intentionality. They create complex impressions with constancy and coherence. In general, Hume’s felt indicator of vivacity has then the function to track mental actions relevant for the creation of an empirical world, one with structured stability. When copied to ideas, it grounds beliefs and demarcates beliefs from mere imaginings. Let us take stock. Hume’s ideas and impressions are systems that are causally coupled by the copy-relation. Copying targets information in the sensory system. This information is based on structure, elements, properties and relations. Much of the structure is based in activities since much of it is not hardwired or innate. Moreover, both systems are mental. Mental activities are not knowable but their effects are. Vivacity is a conscious quality which tracks such effects and profiles of change and it informs about skilful mental activities when specific perceptual outcomes correlate with specific profiles of change. This indicator of skilful mental activity is copied to cognitions along with the rest of the information in the senses. The broad informational matrix ranging from

3. VIVACITY AS ‘QUALIA’ Much of the secondary literature on vivacity fits under the theoretical concept from current philosophy of mind, ‘qualia’. The ‘qualia-interpretation’ of vivacity encompasses three dimensions: ‘folk psychological’, ‘epistemological’ and ‘metaphysical’.6 Let us briefly analyse these and then turn to the secondary literature on vivacity. In the folk psychological sense, qualia refer to the qualitative, ‘felt’ content of conscious states. The milky blueness of the northern sky on a clear spring day would be an example. In its epistemological sense, qualia constitute the ‘sensuous’ foundation of knowledge. The qualia epistemology of empiricists coincides with the Cartesian epistemological ‘marks of the mind’.7 The third dimension concerns the intrinsic nature of qualia.8 Metaphysically, qualia are nonrelational properties. Together, the three 43

9780826443595_Ch02_Final_txt_print.indd 43

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY Empiricist epistemology bottoms out in perceptual theory. The transparent, indubitable epistemology of qualia rests on their metaphysical and phenomenological intrinsicality. This is evident in the jargon used for knowledge of qualia. They are ‘felt’, ‘had’, or ‘acquainted with’ instead of explored and represented. The epistemological puzzles indicate the same. Qualia foundationalism inherits Descartes’s ‘problem(s) of knowledge’ about anything external to inner qualia including the world. Empiricist perceptual theory has a symmetrical problem: senseperceptual consciousness is intentional, about the world, yet qualia are intrinsic, not about anything. The ‘problem of perception’ concerns how sense-perception ‘reaches’ to an external world from intrinsic qualia. An example clarifies this model of the senses and how it influences Hume interpretation. Price describes sense data in the experience of a tomato as ‘a red patch of a round and somewhat bulgy shape’.11 Qualia, sense data, hues and shapes are the given foundation for his empiricist epistemology. But they are also the given for his perceptual theory – ‘something from which all our theories of perception ought to start, however much they may diverge later’.12 For ‘Humean’ empiricists such as Price, intrinsic qualia form the elements of sense-perception; they are the ‘sensations’ of sensory-psychologists. The intentional sense-perception of a tomato is created out of non-intentional hue and shape qualia. How this occurs, however, is a problem in the philosophy of perception. When qualia are identified with sensations of perceptual theory, they are typically also phenomenologically non-representational/intentional states. Vivacity, in the qualia-interpretation, inherits the metaphysics and phenomenology of sensory intrinsicality.13

dimensions form the ‘internalist’ Cartesian picture of the mind: intrinsic qualia are directly known, without inference and representations, and the object of such knowledge is an intrinsic quality, fully internal to the mind and independent of anything external to it. For Hume, vivacity varies in strength across mental state types. ‘Vivacity’ may be taken to refer to the strength in folk psychologically understood qualia. Bennett’s and Stroud’s interpretation of vivacity as ‘phenomenological intensity’ does just that.9 For them, vivacity is Hume’s description of a directly phenomenologically known variation in intensity. For example, if the quale is the colour red, ‘vivacity’ describes its varying intensity so that colour impressions are highly vivid and memory colours less so. The interpretation gains support from Hume’s phenomenological methodology and specific remarks about the uniform nature of vivacity across perceptions: ‘the idea of red we form in the dark and that impression which strikes our eyes in sun-shine differ in degree, not in nature’ (THN 1.1.1.5 / 3). The qualia interpretation does not just concern the folk psychology. Qualia are a close cousin of the epistemological primitives of neo-classical British empiricism. Their ‘sense data’ and ‘the given’ characterize the foundational epistemological role of ‘sensuous’ experience. Price, for example, defines sense data as that which cannot be doubted when having a sensory-experience and identifies them with Humean impressions.10 And since all twentieth-century British neo-classical empiricists are typically seen as Humeans, the qualia-interpretation results in historical continuity. More specifically, as an epistemological qualia-concept, vivacity indicates the foundational role of the senses. Yet common epistemological inclinations do not exhaust the qualia-interpretation. 44

9780826443595_Ch02_Final_txt_print.indd 44

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY gaps in sensory information and to create a belief in an external world.17 And given that beliefs for Hume are phenomenologically vivid ideas, Price’s gap filler ‘hypothesis’ also concerns perceptual phenomenology. Our awareness of a ‘smooth and continuous’ world requires intra-mental higher-order ‘gap fillers’. Contrary to the qualia account, we believe that Hume’s complex impressions have intentional information and that Hume gives a causal explanation of it. His central theoretical sense-perceptual concepts are two conscious qualities, constancy and coherence’ (THN 1.4.2.15–20 / 194–5). Information on the constancy and coherence of sensory objects concerns complex impressions.18 These qualities causally influence imagination to form similar qualities in ideas, ‘continued and distinct existing objects’. Imagination does not ‘supply’ the latter qualities central to the belief-idea in the external world. Instead, imagination processes complex sensory information that already is coherent and constant. Price’s dynamical metaphor of cognitive processes ‘working upon data’ is then misleading. First, it diminishes the information in complex impressions. Second, it overlooks the similarity in information between ideas and impressions, the similarity of the quality pairs, ‘continued and distinct’ and ‘coherent and constant’. Hume relies on this similarity in explanation. According to him, the belief in external objects ‘must arise from a concurrence’ of some ‘qualities’ of ‘certain impressions’ combined with ‘qualities of the imagination’ (THN 1.4.2.15 / 194; emphasis added). In Hume’s full account, idea-intentionality and belief in external bodies is caused by constant and coherent complex impressions. Copying ‘communicates’ causally similar qualities to ideas, viz. distinct and continuous objects.

Sensory vivacity is central for the qualia interpretation so let us focus on it for critical purposes. The qualia model of senses is programmatic to neo-classical empiricism. But is it Hume’s sensory metaphysics? In psychology, as elsewhere, Hume emphasizes the ‘non-relational’ and thus intrinsic aspects of perceptions: ‘every distinct perception, which enters into the composition of the mind, is a distinct existence, and is different, and distinguishable from every other perception’ and that there are no ‘two distinct impressions, which are inseparably conjoin’d’ (THN 1.4.6.16 / 259, 1.2.6.3 / 66). In his theory of senses, Hume emphasizes that simple impressions, colour, taste, smell and touch, ‘admit of no distinction or separation’ (THN 1.1.1.2 / 7–8). These are typically also interpreted as indivisible perceptual ‘atoms or corpuscles’ that make up the idea of space (THN 1.2.3.15 / 30).14 Simple impressions are clearly nonintentional. They lack relations and thus structure, including spatial structure. Because of the intrinsic nature of simple impressions, Hume’s theory of the senses is often seen as the paradigmatic ‘atomistic’ one.15 But are Hume’s complex sensory impressions similarly intrinsic? The common empiricist interpretation of Hume’s account of the origin of clearly intentional perceptions, belief-ideas, shows that complex impressions are not seen to add much to atomic qualia. Price’s classic account of Hume’s theory of the external world, the belief-idea of a separate and independent world, is an example.16 For Price, Hume’s ‘fleeting, perishing impressions’ form a ‘gappy’ sequence of sense data (THN 1.4.2.20 / 195). Even complex impressions give only ‘gappy’ information. Price thinks that higher-order cognitive processes of the imagination supply the needed ‘gap-fillers’. He concludes that Hume’s mind ‘postulates’ to fill in the 45

9780826443595_Ch02_Final_txt_print.indd 45

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY ‘stable experience’. It is created from change in impressions. The causally central facts of change include the variation in stability of complex impressions and the fact that stable ones do not just pop inexplicably into existence. The additional causal posit – vivacity – tracks change and informs about stable, coherent and constant change. Since this type of change is necessary and sufficient for sensory intentionality, vivacity is conscious information on intentionality. And, once copied, vivacity causes the association of ideas with an ‘external world’ instead of internal (causal) fancy. Vivacity is a quality indicative of warranted intentional objects, sense-based ones.22 Let us conclude with a summary. The qualia-interpretation leaves vivacity without an epistemological function. At best, vivacity informs about inner qualitative change. But such change does not inform even about the existence of a world external to the qualia let alone distinguish beliefs from imaginative creations. But for Hume, vivacity informs about the lawlike or constant and coherent change of intentional sensory objects. This function is copied to the idea-realm where vivacity informs about believable ideas through their (intentional) objects’ vivacity. The epistemological emptiness of the qualiainterpretation follows from a mistaken view of Hume’s phenomenology and metaphysics of the senses. Constant and coherent complex impressions are both metaphysically and phenomenologically intentional.

The similarity encompasses the felt intentional object-involving qualities of both perception-types. Principles of the imagination are also involved in the causal process but these cause identity in the intentional objects of ideas. In Hume’s own words, these innate processes ‘render the uniformity as compleat as possible’ once the mind is ‘in the train of observing uniformity among objects’, and the postulation of continued existence ‘suffices’ for this purpose (THN 1.4.2.22 / 197–8). Hume’s account differs significantly from the Cartesian-Kantian nativist approach to cognition. For Hume, internal imaginative processes exaggerate the intentionality of ideas. Ideas are neither a cognitive achievement nor a priori. Moreover, the causes of exaggerated ‘postulation’ extend beyond the mind to language – as they did for Hume’s nominalist predecessors. But whatever their causation, the felt ‘distinct and continued’ objects of ideas – Hume’s pun on Descartes’s distinct and clear idea-objects – are only similar to sensory objects. Since intentional objects and, at times, their causes differ for impressions and ideas, vivacity is also included in Hume’s causal model. Let us see how. Both of the qualities of constancy and coherence pertain to sensory change. For Hume, coherence is directly about the change of complex impressions. When their change profile is lawlike, complex sensory information is ‘coherent’.19 The quality of constancy also concerns change. Hume characterizes it in terms of the ‘order’ of appearance of objects with respect to changes in the perceiver such as his or her head movements.20 This is a mere illustration of the perpetual change in circumstances of perception. Despite these changes, the objects remain constant yet ‘only’ so instead of identical.21 For Hume, lawlike change is the key causal factor in sensory intentionality and

4. VIVACITY AS PRESENTEDNESS We are not alone in underscoring the intentional nature of Hume’s senses. In this section, we analyse Dauer’s intentional interpretation 46

9780826443595_Ch02_Final_txt_print.indd 46

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY aspect of seen colour. We will at times use the term ‘presentational-representedness’ to underscore the intentional nature of Hume’s senses for Dauer. Presentational-representedness causes problems both for the intended phenomenological sense of vivacity and the categorical distinction between impressions and ideas. Bats and humans differ in sensory phenomenology yet their vivid sensory states are about the same intentional object, shape. We will label it ‘shape’ to underscore its phenomenological opaqueness. ‘Shape’ cannot reduce to the ‘transparent’ qualitative shape or colour of human visual experience because bats access ‘shape’ too without experiencing colours. The opaque phenomenology undermines Dauer’s distinction between representational ideas and presentational impressions. For Dauer, representations are many-one ‘mappings’. But both ‘presentations’ and ‘representations’ are in a many-one relationship to a common object. ‘Shapes’ are presented by sounds and colours; ‘shapes’, for Dauer, are represented by equations and sonar. What is the difference between the two many-one mappings? Since presentationally represented objects are opaque in phenomenology, Dauer cannot ground the difference in a phenomenology of the given. The problem applies directly to vivacity. For Dauer, Humean senses do not reduce to qualia. Vivid colours, for example, do not reduce to ‘phenomenologically intense’ colour qualia. Vivacity is an intentional ‘quality’. Dauer also distinguishes presentational-representations from representations in terms of the former’s vivacity. Yet both are intentional. Dauer cannot base this distinction in either qualia or intentionality. As a result, we do not know what role qualia have in an intentional sensory phenomenology, or vice versa. Dauer’s predicament is common for

of vivacity as the ‘what it’s like of consciousness’. For Dauer, vivacity is Hume’s way of describing what it is like to undergo sense-perception, remembering and believing. Although vivacity applies broadly in phenomenology, the core meaning emerges from sensory experience. Dauer defines sensory vivacity in terms of a ‘sense of presentedness’.23 Sensory consciousness presents whereas thinking represents and the two, for Dauer, can ‘never be confused’.24 Thus vivacity demarcates categorically between impressions and ideas and gives the unique phenomenology of sensing. Let us look at the details. Dauer develops the sensory phenomenology of presentedness through the ‘what it’s like’ concept popularized by Nagel.25 Dauer interprets Nagel’s scenario about the different subjective points of view of bats and humans in terms of shape perception. For bats that echolocate, sounds present shapes. That is their vivid experience or ‘what it’s like’ of sensory consciousness. Humans differ subjectively since colours present shapes. For Dauer, it is ‘an essential feature of visual experience’ that ‘colours give us shape’.26 The presentation of ‘shape in colours’ constitutes Humean (visual) vivacity for humans. In Dauer’s account, Humean sensory phenomenology is intentional. It is ‘about’ something. Colour impressions do not just ‘give’ colour as qualia would but ‘give’ shape. Colours are about shapes. Bat phenomenology is also intentional, about shape. Dauer’s presentations of shape occur across different types of sensory experience. This state of affairs, in turn, is ‘definitional’ for intentionality. Intentionality is ‘aspectual’: intentional states reach out to their objects ‘under this aspect rather than another’.27 Bats’ sensory states reach out to the ‘object’ of shape under the aspect of heard sound; humans’ sensory states reach out to that object under the 47

9780826443595_Ch02_Final_txt_print.indd 47

2/9/2012 9:39:34 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY play a role in intentional sense experience if changing colour impressions have a role in the creation of constant and coherent complex impressions. It is very likely that colours have a systematic role in this. Sensed objects have systematic colour-profiles of change. Such ‘coherent and constant’ change concerns change in the colour of objects relative to our action and changes in the objects and their perceptual environment.29 For Hume, colours have a role in ‘presentedness’ of sense-perception because they have a role in the causation of intentional sensory objects. What is the role and nature of vivacity for intentionality? Vivacity is an indicator of change and lawlike profiles of change make intentional sensory objects in the first place. Vivacity has a central causal role. But what is the phenomenological nature of vivacity relative to philosophical terms of art such as the ‘given’? Dauer equates vivacity with a given sense-perceptual presentedness. But liveliness cannot characterize momentary experiences, states or ‘perceptual objects’. Vivacity pertains to change and characterizes events. If ‘the given’ excludes definitionally the phenomenology of events, vivacity cannot be given. But to understand Hume we must find room for an event phenomenology. The conscious qualities of constancy and coherence of complex impressions themselves characterize change. Intentional sensory objects are created from such change and are always in a process of making. Nothing in an experiential state corresponds to the quality of vivacity; nothing in experience corresponds to a fixed sense-perceptual object. As Noë, the enactivist perceptual theorist, puts it, there are no ‘snapshot pictures’ in experience.30 Our Hume would agree. Still, ‘presentedness’ does capture a fundamental feature of sensory experience.

intentional accounts of sense-perception. The relationship, if any, between the phenomenologically given and representational object/ content is the core puzzle for such accounts. Without an answer to it, colours could be seen to ‘give’ shape non-phenomenologically. Indicator-theories of sensory content do just that. Colours serve the function of shape perception without being ‘felt’.28 But a merely ‘indicated’ intentional object is not a given in phenomenology just as a represented object need not be felt in its representation. Both are opaque representations. Is there a way to resolve Dauer’s problem in order to advance a representationalintentional account of Humean senses that includes vivacity? Dauer’s predicament rests on the specificity of the representational object, ‘shape’, and representer, colour, together with the opaqueness of phenomenology. This potentially initiates a problematic slide towards indicator theories. ‘Shape’ could though refer to the felt world-involving feature of sensory experience as such. If so, ‘shape’ is neither a specific type of quale nor an opaquely indicated domain. Instead, ‘shape’ refers to the intentional ‘feeling’ of vision. If this is Dauer’s view of Humean vivacity, it is promising, although it is not an accurate account of Hume’s view. Let us see how colours might play a role in vivid or intentional sensory experience according to Hume’s theory of senses. Colours qua simple impressions do not ‘give us’ anything in phenomenology. Simple impressions are non-intentional qualia. Not even all Humean ‘coloured’ complex impressions ‘give’ us anything. For example, a still and unchanging or ‘uniform’ colour mosaic will lack intentional appearance. This is because, for Hume, only complex impressions with felt coherent and constant change-profiles appear intentional. Colours 48

9780826443595_Ch02_Final_txt_print.indd 48

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY behind intentional objects overlap only for ‘sensing and thinking’. Contentful ideas have many causes. In epistemology, Hume is concerned with errant causes, imaginings and linguistic habits, and their fictitious idea-effects. Equating believability with an indicator copied from the senses addresses this concern. Whether vivacity should be seen as a ‘given’ according to the conceptual habits of some later empiricist movements is far less important than knowing what vivacity teaches us. For Hume, it informs us of the sensory origin of some representational objects and teaches us which ideas are worthy of belief.

To articulate this ‘given’, we must bring into view the change-oriented nature of the presented world. Noë refers to it as a ‘felt presence’.31 Sensory awareness of a world concerns the availability of further detail about its objects. We feel this and this is because of change in the sensed objects and their circumstances. Our Hume would agree. He would underscore that the ‘felt presence’ is a ‘given’ only once we have learnt the indicator function of vivacity with respect to profiles of change in sensory objects. We are accustomed to ‘just feel’ how the sensed world will change according to learnt profiles of change. We learn associations and feel them. A better word for this phenomenology is a felt ‘giving’. As a ‘given’, vivacity must be seen in intentional terms, as being about the further information available to us. That is what vivacity tracks and what it has taught our sensory phenomenology.32 Our account of ‘the giving’ in vivacity applies to ideas in Hume’s comparative sense. The intentional objects of ‘idea-events’ change less and for several reasons. Cognition is ‘off-line’ and not open to external change in the manner ‘on-line’ sense-perception is. Sense-perception allows us to access information from the world instead of just (copied) memories. Moreover, ideas traffic in ‘fictitious’ categories based in the philosophical relation of identity. Extra-sensory, intramental processes ‘reify’ their intentional objects in terms of identity, whose effects are not felt in the senses – as each of us can verify by attending carefully to the perpetual change of sensory objects. For these reasons, ideas lose in vivacity to the ‘constant and coherent’ intentional sensory objects. Hume’s epistemological message with vivacity is that a fully empirical, impressionrelated indicator applies to ideas. Hume’s empiricism is externalist. The causal processes

5. VIVACITY AS VERISIMILITUDE In this section, we analyse another intentional interpretation of vivacity, Waxman’s ‘verisimilitude’. The term was originally used by Popperians in philosophy of science to refer to the ‘truth-likeness’ of scientific theories. But Waxman’s ‘verisimilitude’ is not a concept of metaphysics or even an epistemological judgement of representations. Waxman uses it to describe the ‘truth-likeness’ of Humean perceptions. For Waxman, ‘before there can be belief in the existence of any perception, consciousness must respond to it with the feeling that it is real’.33 Verisimilitude is this phenomenological ‘reality-likeness’ of perceptions, their felt ‘presence and reality’.34 Let us first explore Waxman’s idea by comparing it with the two earlier interpretations and then turn to a critical analysis. Waxman’s verisimilitude reading differs from the qualia-interpretation. Verisimilar perceptions are a response to qualia, whereby the intrinsic states become about reality. Waxman’s interpretation of vivacity is similar to Dauer’s ‘presentedness’. Yet Waxman 49

9780826443595_Ch02_Final_txt_print.indd 49

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY something ‘attending perceptions’.39 Waxman concludes that vivacity is an ‘original division’ of consciousness for Hume.40 Waxman’s analysis of vivacity incorporates the causal history of the relevant phenomenology. Waxman’s Hume is a nativist: ‘one could in principle distinguish impressions and ideas from the first moment of one’s conscious life.’41 This nativism is based in an innate vivacity. Nativism entails that intentionality is not caused by sensory experience or intramental cognitive processes such as judgement. Or more generally, since vivacity is not a quality of perceptual objects, it cannot be affected by the causes that affect Humean perceptual objects. Nativism also influences Waxman’s phenomenology. Vivacity is an intentional ‘frame’ of phenomenology and cannot vary across specific types of intentional objects. Let us turn to critical analysis. Waxman pairs nativism with naturalism. By leaving verisimilitude with ‘human nature’ and by conceptualizing it as having a ‘primitive character’ that ‘applies to already animals and infants’, Waxman intends to do justice to Hume’s naturalism.42 But, historically, nativism is associated with the anti-naturalism of the rationalists. Not surprisingly, Waxman’s nativist reading contradicts Hume’s causal theory of intentionality. Let us briefly revisit Hume’s account to see why. Waxman ignores the distinction between simple and complex impressions and their qualities. Simple impressions have no intrasystemic causal history for Hume. If we access these in phenomenology, they are qualia. Simple impressions are ‘original’ with respect to Hume’s theoretical psychology and maybe to his account of consciousness. Neither is true of all complex impressions. First, all complex impressions have a causal history in processing. Much of this falls below the threshold

overcomes the ambiguity about the object that is presented in phenomenology and the un-Humean narrow applicability of presentedness only to impressions. Waxman’s account underscores the general worldinvolving nature of vivacity for the phenomenology of both ideas and impressions. He distinguishes vivacity from Humean ‘perceptual objects’. The latter are mere qualia or ‘appearances’.35 Verisimilitude is not their quality (of intentionality) but a ‘quality of consciousness’. As a result, Waxman would disagree with Dauer’s contention that, in visual phenomenology, colours ‘give’ or ‘present’ us with shape. Colours as perceptual objects are qualia that present or ‘give’ nothing. Once colours are verisimilar, they are intentional but not about any specific object (such as shape) nor opaquely so (as they would be if they were about ‘shape’). Verisimilar visual experience presents a full-blown, intentional, coloured reality. To achieve interpretative overlap and conceptual coherence, it is best to analyse Waxman’s distinction between qualities of consciousness and perceptual objects as two different ‘styles’ of consciousness. In one, perceptions are merely ‘present’ as qualia to consciousness.36 For Waxman, this style of consciousness is not vivid. The other style is. Consciousness ‘responds’ to idea and impression qualia with a feeling of reality or verisimilitude. These perceptual objects are intentional. For Waxman, both styles of consciousness are a matter of ‘human nature’ or ‘reflect its constitution’.37 In other words, both are innate. The terms Waxman uses for the role of consciousness in imbuing Humean qualia with verisimilitude – ‘act of the mind’, a ‘regarding’, a ‘stance’ – are then misleading in their causal, process perspective.38 All Humean perceptual objects acquire vivacity just ‘in’ consciousness, so vivacity is 50

9780826443595_Ch02_Final_txt_print.indd 50

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY of consciousness with respect to ‘felt’ indicators. To the extent that it does, the causal origins of complexes have no corresponding original divisions in ‘qualities’ of consciousness. But this is not true of the causation of constant and coherent complex impressions. Such structures are caused by constant and coherent change. It is reflected in consciousness through effects, constant and coherent sensory objects and vivacity. Furthermore, these qualities, together with copying, cause the belief in a separate and distinct external world. And this idea, in turn, has precisely the phenomenological identity Waxman refers to with ‘verisimilitude’, the felt presence of an external reality. As a consequence, neither verisimilar ideas nor verisimilar complex impressions are phenomenologically ‘original’ or causally innate. Both have a causal history that extends to our consciousness of them. And when we add that Hume’s account also concerns specific intentional objects and not just the intentional ‘frame’ of phenomenology, we see that Waxman’s interpretation misses its mark also in the role of learning. The indicator function of sensory vivacity with respect to specific coherent and constant objects is learnt. Only through a successful causal learning-history can copied vivacity ground the epistemology of believability within ideas. Epistemologically warranted, verisimilar consciousness is a learnt style of consciousness without an influence from unnaturalistic ‘judgemental’ processes.

The heuristic leaves the functionalist interpretation straddling the fence between externalism and internalism. Since our own view underscores mental functioning, a few remarks on it are in order to bring our analysis of the secondary literature to a close. For functionalists, vivacity concerns the system-wide significance of perceptions, their ‘force’ in influencing other perceptions.44 Vivacity is the intramental ‘functional role’ of perceptions. And such roles vary since perceptions differ in their degree of interconnectedness and their associated felt ‘force’. Functionalists have an explanation of Hume’s identification of belief with vivacity. Believable ideas have interconnections and a felt influence on other ideas, whereas mere ideas do not. Sensory vivacity is problematic for functionalists and especially when impressions are seen as non-relational qualia. In her seminal functionalist interpretation, Govier solves the problem by dividing Hume’s vivacity-terms into two categories.45 ‘Forceful’ and ‘vigorous’ are functional terms. ‘Vivid’ and ‘lively’ are not but describe the clarity and amount of detail of a perception. Govier argues that the functional concept applies to ideas and that sensory vivacity concerns only clarity and detail. Govier’s Hume turns out to be a coherentist instead of an empiricist. Vivacity of beliefs has nothing to do with vivid impressions. Copied sensory vivacity concerns only ‘informational quality’ – clarity and detail – which is without epistemological significance for Govier. Hume’s epistemology rests fully on the interconnections among ideas. This is a distorted picture. Yet Govier’s interpretation can be diagnosed in a manner that supports Hume’s central epistemological idea, the relationship between sensory information and cognitive function.

6. VIVACITY AS FUNCTIONAL ROLE Vivacity is a copied epistemological indicator. In accordance with this idea, we have ignored interpretations of vivacity that emphasize the internal processing of perceptions.43 51

9780826443595_Ch02_Final_txt_print.indd 51

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY becomes a logical extension of his ‘genetic’ psychology of sensory intentionality. But coherence can make myths. A historical hypothesis must also be fruitful. Our concluding remarks are intended to achieve just that, offer ideas for a reassessment of Hume’s place in the history of philosophy. ‘Idea’ is the epoch-making notion of preKantian, ‘early-modern’ philosophy. Hume’s philosophy is one theory of ideas. History has not dealt kindly with the ‘idea-idea’. Linguistic and/or logically structured propositions replaced ideas without a trace from the analytic tools of epistemologists. Ordinary languages, forms of life, and other holistic epistemologies buried ideas under mangled practices. ‘Ideas’ are so thoroughly discredited that addressing Hume’s version appears of antiquarian interest, at best. But the current wisdom is itself importantly historical: critics of early empiricism learnt their early empiricism from much later empiricists. If only for this reason, Hume’s ‘ideas’ deserve a fresh look. Hume’s empiricist predecessors never recovered ideas from the Cartesian mental depths. But Hume did and by meeting Descartes on his turf. Hume too articulated an ‘inner methodology’ for ideas. Cartesian clear and distinct ideas were replaced with vivid ideas. This indicator of believable ideas was copied from the senses. Hume’s externalist epistemological internalism was usable. Fictitious ideas including dreams are felt to be unbelievable, phenomenologically, through idiosyncracies in vivacity. Fictions can be identified ‘internally’, with vivacity of copied memories and idea-acts – without further experience. Hume thought that the inner methodology could even fight ‘custom’ and its imaginative necessary connections.47 The ‘internal arm’ of Hume’s empiricism was decisive on the ‘problem of the external

Only intentional complex impressions have Govier’s informational qualities of detail and clarity. Only a specific type of change in impressions, a specific ‘functional role’, causes impressions with both detail and clarity. The phenomenology of ‘felt presence’ or ‘giving’ shows that sensory objects take clear form through detail. Sensory information integrates functional profiles with informational quality. This allows for an inner epistemology for ideas: through copying, detail and clarity become criteria for the intentional objects of ideas which themselves are an effect not just of copying but also of ‘independent’ intramental cognitive functioning. The skill of thinking and internal functioning is itself governed by skilful sensory functioning through shared criteria for the intentional objects of both. Vivacity is a ‘functional role’ indicator but an empirical one, without a trace of systemwide coherentist epistemological functions. Unintentionally, Govier gives evidence in favour of our integrated interpretation.46 In seventeenth-century English, her term for clarity and detail, ‘vivid’, meant ‘lively’, her term for functional role. Perhaps Hume’s use of ‘vivid’ caused a natural association between ideas of change and stability in the minds of his audience?

7. CONCLUDING REMARKS: HUME’S DEVELOPMENT OF THE THEORY OF IDEAS Our interpretation has followed a traditional internalist historiography. Its goal is coherence. Hume’s theory becomes coherent when vivacity is understood as a phenomenological indicator of an ‘inner methodology’ copied from the senses; Hume’s epistemology 52

9780826443595_Ch02_Final_txt_print.indd 52

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY Cartesian internal depths even if he stopped short of turning them into logical, social artefacts. But did he drag ideas to the world? His perceptions constitute an ‘empirical world’. But this is not based in transcendental reasons of a priori cognitive achievements. The innate a priori categories of Kant’s nativistempiricism ‘schematized’ in the senses are not a limit for or origin of Hume’s empirical world. It is a labile causal interface of perception and action. Complex structure is created from change and its only resting place is coherent and constant change accessed from the world, as such, unknowable. Hume was then an empiricist about the nature of empiricism, unlike Kant and the logical positivists. In this ‘naturalism’, Hume is reminiscent of today’s epistemologists who deny ‘first philosophy’. He never gave a ‘first version’ of his empiricism either. Yet the modern naturalism is based in Quine’s view of holistic knowledge, one that ignores psychology and the knower’s awareness. Modern naturalism is eliminativist in its externalism: a knower need not know that he or she knows to know. Justification exists below and beyond the radar of the knower, in phylogenetic processes (evolutionary epistemology), in the social epistemology of scientific practices (Quine and Putnam’s division of linguistic labour), and unconscious neural processes (Churchlands’s eliminativism). Each demands that the agent defers justification to external forces and loses touch with the felt dimension of knowledge to know. In Hume’s naturalism, only the phenomenological afterglow of associations can create justified beliefs and an epistemic agent. He never let go of the role of the subject’s learnt skilful awareness for knowledge. In its developmental, individualistic and phenomenological orientations, Hume’s account is

world’ and with facts closest to home, consciousness of change in perceptions. Despite their overlap, Hume did not extend Descartes’s project. To Hume’s detriment, it was Descartes’s project that through neo-Kantians won a decisive (interim) victory in the competition over the sociological aspects of epistemology. Cartesian-Kantian epistemology focuses on ‘form’, representational frameworks. For understanding Hume, the differences between the form of idealized languages and clear and distinct ideas is irrelevant. Hume had no epistemology of form because he believed no representational framework existed nor was required for the mind to know. Abstract ideas are fictions because no abstract mental representational vehicle exists. Without one, information cannot be structured by identity and difference, whether codified in clear and distinct ideas or analytic definitions. His disregard of form entails that the textbook image of Hume as the forefather of logical positivism is misleading. His structured knowledge is inexpressible through definitions, whether sensation, observation or operation language-based. The reason is that his knowledge-mechanism does not follow a syntax or grammar. Thus his empiricism does not rest on empirical atomic propositions – Tractarian or Russellian logical atoms that mirror empirical facts – that turn into complex propositions by truth-functional connectives. Hume’s mechanism is not rulegoverned at all because it is causal; it is governed by all of the causally robust dimensions of acts of experience. No logic governs these causal principles; complex information is a contingent matter. Nor can it be studied in the abstract or normatively as grammar and logic are. In short and unlike that of the logical positivist, Hume’s empiricism is naturalistic. Still, he did drag ‘ideas’ from the 53

9780826443595_Ch02_Final_txt_print.indd 53

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY too limited to solve many of today’s burning epistemological problems. The contrast in naturalisms appears to us then as much historical as it is substantive. Historically, Hume’s epistemology is of and for the Enlightenment. From the Enlightenment perspective, current naturalism is not only psychologically implausible but socially irresponsible. Substantially, if the Enlightenment philosophers tackled problems still relevant today, Hume offers a systematic naturalistic empiricism for addressing them. In both cases, our reassessment of his copying should help to see the history of philosophy and especially empiricism in a novel, fruitful manner.

NOTES 1

2

3

4

A comprehensive analysis would also cover Hume’s other quality, ‘facility’ (THN 1.3.13.6 / 146; 2.3.5.5 / 422; EHU 4.2 / 25). We will not explore facility on this occasion, but merely remark that facility is clearly dependent on vivacity since no ease of transition within perceptions can occur without identification of transitions or patterns of change. See P. Godfrey-Smith, Complexity and the Function of Mind in Nature (Cambridge: Cambridge University Press, 1996), chap. 2. See H.H. Price, ‘The Permanent Significance of Hume’s Philosophy’, in A. Sesonske and N. Fleming (eds), Human Understanding: Studies in the Philosophy of David Hume (Wadsworth Publishing, 1968), p. 7. See also A. Rosenberg, ‘Hume’s Philosophy of Science’, in D.F. Norton (ed.), The Cambridge Companion to Hume, 2nd edn (Cambridge: Cambridge University Press, 1993), p. 66; G. Dicker, Hume’s Epistemology and Metaphysics: An Introduction (London: Routledge, 1998), p. 11ff.; J. Bennett, Locke, Berkeley and Hume: Central Themes (Oxford: Clarendon Press, 1971), chap. 9; Learning from Six Philosophers, vol. 2 (Oxford: Oxford University Press, 2001), chap. 32; and

5

6

7

8

‘Empiricism about Meanings’, in P. Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), chap. 3; A. Flew, David Hume: Philosopher of Moral Science (Oxford: Blackwell, 1986), pp. 22–3 and D. Pears, Hume’s System (Oxford: Oxford University Press, 1990), pt 1. More recent advocates of a semantic interpretation of copying focus on ‘concepts’, stored mental representations that are used in thinking or cognition more broadly (see, for example, J. Fodor, Hume Variations (Oxford: Clarendon Press, 2003). Although this idea has important differences from the other one, we will not differentiate the two on this occasion. The main reason is that the overall outcome for the ‘fate’ of copying has been largely the same here as it has in the more directly linguistically-oriented semantic traditions. For an exception, however, see J. Prinz, Furnishing the Mind: Concepts and Their Perceptual Basis (Cambridge, MA: MIT Press, 2002), chap. 5. For an overview of the main puzzlements that surround Hume on belief see M. Gorman, ‘Hume’s Theory of Belief’, Hume Studies 19(1) (1993), pp. 89–102. No doubt the fuel to this scholarly fire is provided by Hume’s own linguistic dissatisfaction with the account of belief in the Appendix to the Treatise, admitting that he is at a ‘loss for terms to express [his] meaning’ (THN 1.3.7.7 / 628). R. Descartes, Rules for the Direction of the Mind [1628] in J. Cottingham et al. (eds), The Philosophical Writings of Descartes, vol. 1 (Cambridge: Cambridge University Press, 1984–91). These three dimensions map onto Dennett’s influential analysis although he does not label them; see ‘Quining Qualia’ in W. Lycan (ed.) Mind and Cognition: A Reader (Cambridge, MA: MIT Press, 1990), pp. 519–47. Originally in A. Marcel and E. Bisiach (eds), Consciousness in Modern Science (Oxford: Oxford University Press, 1988). Dennett includes Cartesian marks such as privacy and direct knowledge in ‘Quining Qualia’, p. 519 (see also, e.g., J. Kim, Philosophy of Mind, 2nd edn (Boulder, CO: Westview, 2006), pp. 18–22). Dennett lists other key ideas used to characterize the ‘intrinsicality’ of qualia: intrinsic

54

9780826443595_Ch02_Final_txt_print.indd 54

2/9/2012 9:39:35 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY

9

10

11 12 13

14

15

16

17

18

qualia are non-relational properties – dynamically – they do not change depending on the experience’s relation to other things, and, synonymously or interchangeably, intrinsic properties are ‘simple’, ‘homogenous’, ‘unanalyzable’ and/or ‘atomic’ (‘Quining Qualia’, p. 519). See Bennett’s Locke, Berkeley and Hume, p. 225; Stroud’s Hume (London: Routledge & Kegan Paul, 1977), p. 29; J. Broughton, ‘Impressions and Ideas’ in S. Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), p. 45; Dicker’s Hume’s Metaphysics, p. 5 and D. Owen, Hume’s Reason (Oxford: Oxford University Press, 1999), p. 73. See H.H. Price, Perception (London: Methuen & Co., 1932), pp. 3, 19. See also Bennett’s Locke, Berkeley and Hume, p. 222 and Flew’s David Hume, p. 26. Price, Perception, p. 3. Ibid., p. 19. Since vivacity attaches to all Humean perceptions, according to the qualia-interpretation, all mental states have a non-representational dimension since all of them have a qualia dimension of vivacity. D. Raynor, ‘Minima Sensibilia in Berkeley and Hume’, Dialogue 19 (1980), pp. 196–200; D. Garrett, Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997), pp. 60–2; and W. Waxman, Hume’s Theory of Consciousness (Cambridge: Cambridge University Press, 1994), pp. 44–6. T. Holden, ‘Infinite Divisibility and Actual Parts’, Hume Studies 28(1) (2002), pp. 3–26, p. 4; M. Frasca-Spada, ‘Simple Perceptions in Hume’s Treatise’, in E. Mazza and E. Ronchetti (eds), New Essays on David Hume (Milan: FrancoAngeli, 2007), p. 42. H.H. Price, Hume’s Theory of the External World (Oxford: Clarendon Press, 1940). Ibid., chap. 2. For other qualia accounts on this score see Stroud’s Hume, chap. 5 and Dicker’s Hume’s Metaphysics, p. 169. Constancy and coherence cannot be qualities of simple impressions for they are ‘fleeting’, ‘perishing existences and appear as such’ (THN 1.4.2.15 / 195). That these qualities are of complex impressions is clear from Hume’s examples such as mountains, houses and trees,

19

20

21

22

or a bed, table, ‘books and papers’, ‘sun or ocean’. Hume also gives the scenario where he leaves the room for an hour and finds when he comes back that the fire is not in the same way as he left it (THN 1.4.2.18–19 / 194–5). These illustrations all involve complex impressions with clear intentional objects even if processes. Re-entering the room he finds the fire not in the same way as he left it, but he is accustomed in other instances to see a ‘like alteration produced in a like time, whether [he] is absent, present, near or remote’ so this coherence of ‘their changes is one of the characteristics of external objects’ (THN 1.4.2.19 / 195). In the case of constancy he writes: ‘Those mountains, and houses, and trees, which lie at present under my eye, have always appear’d to me in the same order; and when I lose sight of them by shutting my eyes or turning my head, I soon after find them return upon me without the least alteration. My bed and table, my books and papers, present themselves in the same uniform manner, and change not upon account of any interruption in my seeing or perceiving them’ (THN 1.4.2.18 / 195). See, for example, A. Noë, Action in Perception (Cambridge, MA: MIT Press, 2004). The idea that vivacity informs about stable complex impressions creates conceptual uniformity among the many alternative terms Hume uses for ‘vivacity’ such as strength, violence, liveliness, vigour, firmness, steadiness and solidity. From an ordinary language point of view, the list contains unrelated and even contradictory terms. For example, ‘solidity’ seems unrelated if not antagonistic to ‘liveliness’. Solid things such as statues are inert, unmoving and generally lacking in liveliness. But from our interpretative perspective, the two terms are naturally connected. ‘Solid’ complex impressions are ones whose intentional objects remain constant despite change and because they display a robust ‘lively’ profile of change. The object remains stable amidst change by emerging from lawlike change. The solid objects are also found out to be so through our own action which, from a sensory end point, consists of ‘liveliness’ or patterned change in impressions. In short, solidity and ‘firmness’ emerge from liveliness or vivacity in the same way that constancy amidst change

55

9780826443595_Ch02_Final_txt_print.indd 55

2/9/2012 9:39:36 PM

HUME’S EMPIRICIST INNER EPISTEMOLOGY

23

24 25

26 27

28

29

30

31

and coherence of change complement each other. We leave it to the reader to work out how other apparently unrelated if not antagonistic terms Hume uses to characterize vivacity indicate the dynamic between stability and change. That Hume’s apparently unrelated linguistic descriptors form a coherent whole lends support to our interpretative perspective. F.W. Dauer, ‘Force and Vivacity in the Treatise and the Enquiry’, Hume Studies 25(1&2) (1999), pp. 83–100, p. 86. Ibid. T. Nagel, ‘What Is It Like to Be a Bat?’, Philosophical Review (1974), pp. 435–50. Dauer, ‘Force and Vivacity’, p. 86. G. Graham, Philosophy of Mind (Cambridge: Cambridge University Press, 1993), p. 136. In indicator theories, sensory content is identified with representational content. It has no phenomenological ‘intrinsic’ features. In some versions, we just ‘see through’ experience to its representational content instead of accessing qualia (e.g. G. Harman, ‘The Intrinsic Quality of Experience’, in Philosophical Perspectives 4 (1990), pp. 31–52). In Dretske’s version, colours can represent surface spectral reflectance without it being transparent in phenomenology (except in colour constancy); see F. Dretske, Naturalizing the Mind (Cambridge, MA: MIT Press, 1995). We can ‘see’ nothing about the object represented but still our content co-varies with and provides information on it. All versions are problematic both for Dauer’s idea of a phenomenologically given vivacity and the distinction between impressions and ideas in terms of intentionality or vivacity. For details, see, for example, Noë’s Action in Perception and J. Broackes, ‘The Autonomy of Colour’, in K. Lennon and D. Charles (eds), Reduction, Explanation, and Realism (Oxford: Oxford University Press, 1992), pp. 421–65. See Noë’s Action in Perception, p. 35.

32

33

34 35 36 37

38

39 40 41 42

43

44

45

46 47

Ibid. Hume’s distinction between present in fact and present in power can be easily used to articulate this phenomenology and the dynamic between change and stability. Hume does so himself for ideas; see THN 1.1.7.7 / 20. Waxman supports the verisimilitude reading from passages where Hume talks about ‘acts of the mind’ that have a reality-revealing phenomenological indicator quality; see Hume’s Theory of Consciousness, p. 33ff. Ibid., p. 34. Ibid., pp. 38–9. Ibid., p. 34. Ibid., p. 35 and also Waxman’s ‘Impressions and Ideas: Vivacity as Verisimilitude’, Hume Studies 19(1) (1993), pp. 75–88, p. 77. Waxman, Hume’s Theory of Consciousness, pp. 33–5, 38, and ‘Impressions and Ideas’, pp. 77, 81. Waxman, Hume’s Theory of Consciousness, p. 42. Ibid., p. 27. Waxman, ‘Impressions and Ideas’, p. 75. Waxman, Hume’s Theory of Consciousness, pp. 38, 42. For example, Stevenson’s account of consciousness as an impression of reflection in ‘Humean Self-consciousness Explained’, Hume Studies 24 (1998), pp. 95–130. See, for example, S. Everson, ‘The Difference between Feeling and Thinking’, Mind 97, (1988), pp. 401–13. T. Govier, ‘Variations on Force and Vivacity in Hume’, Philosophical Quarterly 22 (1972), pp. 44–52, p. 45ff. Govier, ‘Variations’, p. 46. For learning contexts geared towards novelty and general applicability, such as those of science, the inner arm had to be refined with a codified and more rigorous external arm, in essence, Baconian or ‘Millian’ methods of causal inference (THN 1.3.15 / 173ff.).

56

9780826443595_Ch02_Final_txt_print.indd 56

2/9/2012 9:39:36 PM

3 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION Peter Millican

‘Is Hume a sceptic about induction?’ This might seem to be a fairly straightforward question, but its appearance is misleading, and the appropriate response is not to give a direct answer, but instead to move to a more fundamental question which is suggested by Hume himself at the beginning of his definitive discussion of scepticism in Enquiry Section 12: ‘What is meant by a sceptic?’ (EHU 12.2 / 159). His point here is that ‘sceptic’ can mean many things, and what counts as ‘sceptical’ will often depend on the relevant contrast. Someone who is sceptical about morality or the existence of God, for example, need not be sceptical about the external world. And someone who is sceptical about the rational basis of inductive inference need not be sceptical at all – in the sense of dismissive or critical – about the practice itself. This crucial point about the varieties of scepticism is often overlooked in discussions of Hume on induction, generating a great deal of misunderstanding. Commonly, the debate will be framed in terms of a simple contest between ‘sceptical’ and ‘non-sceptical’ interpretations. Then on the one side, a case is made drawing on Hume’s famous negative argument which apparently denies induction any basis in ‘reason’.1 Meanwhile,

on the other side, appeal is made to Hume’s writings as a whole – including the Treatise, Essays, Enquiries, Dissertations, History and Dialogues – which display a clear commitment to induction, and even reveal their author to be a fervent advocate of inductive science. The evidence on each side is then judiciously weighed, and an appropriate conclusion drawn depending on which way the balance falls. But this whole procedure is misdirected, because once we recognize the varieties of scepticism, it becomes clear that these two bodies of evidence are not in conflict.

1. A SCEPTICAL ARGUMENT, WITH A NON-SCEPTICAL OUTCOME In this chapter, I shall maintain that Hume’s argument concerning induction is indeed a sceptical argument, in the sense of showing that inductive extrapolation from observed to unobserved lacks any independent rational warrant. To avoid any misunderstanding on the way, however, it will help to be clear from the start that this is entirely compatible with his wholehearted endorsement of such extrapolation as the only legitimate method 57

9780826443595_Ch03_Final_txt_print.indd 57

2/9/2012 9:38:48 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION then we have no good reason for supposing that human life will indeed perish in these circumstances. But Hume suggests that even the Pyrrhonist – whatever his theoretical commitments – will be quite unable to insulate himself from such common-sense beliefs: ‘Nature is always too strong for principle. . . . the first and most trivial event in life will put to flight all his doubts and scruples, and leave him the same, in every point of action and speculation’ with the rest of us (EHU 12.23 / 160). Hume cannot, of course, prove that putting total scepticism into practice will lead inevitably to disaster, at least not to the satisfaction of the Pyrrhonist who consistently refrains from induction. Nor can he prove that common life will always trump sceptical principle. But if in fact Hume’s inductive conclusions about human psychology are correct, then he does not need to prove these points to any such opponent:

for reaching conclusions about ‘any matter of fact, which lies beyond the testimony of sense or memory’ (EHU 12.22 / 159). The two may initially seem incompatible, but if so, this is because we are taking for granted that a method of inference is to be relied upon only if it can be given an independent rational warrant. And one of the central messages of Hume’s philosophy is that this assumption is itself a rationalist prejudice that we should discard, even though it is shared by both the Cartesian dogmatist and the extreme ‘Pyrrhonian’ sceptic. In the contest between those two extremes, the Pyrrhonist ‘seems to have ample matter of triumph’ while he ‘justly’ urges Hume’s own ‘sceptical doubts’ of Enquiry 4 (the famous argument which is then summarized at EHU 12.22 / 159). However, the appropriate response, as Hume himself explains, is not to follow the dogmatist in vainly attempting to challenge the argument that yields these doubts, but rather to ask the Pyrrhonist: ‘What his meaning is? And what he proposes by all these curious researches?’ (EHU 12.23 / 159). What, after all, does he really expect us to do in response to this sceptical argument, even if we fully accept it? Is he seriously proposing that we should stop drawing inferences about the unobserved? That would obviously be absurd:

Nature, by an absolute and uncontroulable necessity has determin’d us to judge as well as to breathe and feel; nor can we any more forbear [making inductive inferences], than we can hinder ourselves from thinking as long as we are awake, or seeing the surrounding bodies, when we turn our eyes towards them in broad sunshine. Whoever has taken the pains to refute the cavils of this total scepticism, has really disputed without an antagonist, and endeavour’d by arguments to establish a faculty, which nature has antecedently implanted in the mind, and render’d unavoidable. (THN 1.4.1.7 / 183)

a Pyrrhonian . . . must acknowledge, if he will acknowledge any thing, that all human life must perish, were his principles universally and steadily to prevail. All discourse, all action would immediately cease; and men remain in a total lethargy, till the necessities of nature, unsatisfied, put an end to their miserable existence. (EHU 12.23 / 160)

So if in fact the sceptic’s doubts will be spontaneously ‘put to flight’ as soon as common life intrudes, then Hume’s point is practically successful even if theoretically unproved. And recall again that Hume himself need not be committed to accepting only what is

Theoretically, the Pyrrhonist might try to deny any such disastrous consequences, on the ground that if induction is unwarranted, 58

9780826443595_Ch03_Final_txt_print.indd 58

2/9/2012 9:38:49 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION Such antecedent scepticism is utterly unworkable, because in refusing to trust our faculties from the start, we are denying ourselves the only tools that could possibly provide any solution. The proper alternative, Hume seems to be saying, is to accord our faculties some initial default authority, and to resort to practical scepticism about them only ‘consequent to science and enquiry’, in the event that those investigations reveal their ‘fallaciousness’ or ‘unfitness’ (EHU 12.5 / 150). Thus the onus is shifted onto the sceptic to give reasons for mistrusting our faculties, and in the case of induction, that onus is at most only partially fulfilled. Admittedly,

theoretically provable – that is the very prejudice which he is aiming to undermine. Hume’s subtle approach to scepticism is made harder to appreciate by the vigour and rhetoric of some of his negative arguments and conclusions (especially in the Treatise, where his ultimate position on scepticism remains relatively obscure), but also, I suspect, by the widespread tradition of approaching scepticism initially through Descartes’s Meditations. Descartes sees the sceptic as an opponent to be refuted outright, through rational argument of such overwhelming force as to be immune to any possible doubt. He thus takes on the onus of providing an ultimate justification of human reason, with any ineradicable doubt telling in favour of his sceptical opponent. Hume succinctly points out the fundamental flaw in this approach immediately after having raised the question ‘What is meant by a sceptic?’ at the beginning of Enquiry Section 12:

The sceptic . . . seems to have ample matter of triumph; while he justly insists, that all our evidence for any matter of fact, which lies beyond the testimony of sense or memory, is derived entirely from the relation of cause and effect; that we have no other idea of this relation than that of two objects, which have been frequently conjoined together;2 that we have no argument to convince us, that objects, which have, in our experience, been frequently conjoined, will likewise, in other instances, be conjoined in the same manner; and that nothing leads us to this inference but custom or a certain instinct of our nature; which it is indeed difficult to resist, but which, like other instincts, may be fallacious and deceitful. (EHU 12.22 / 159)

There is a species of scepticism, antecedent to all study and philosophy, which is much inculcated by Des Cartes . . . It recommends an universal doubt . . . of our very faculties; of whose veracity . . . we must assure ourselves, by a chain of reasoning, deduced from some original principle, which cannot possibly be fallacious or deceitful. But neither is there any such original principle, which has a prerogative above others, that are self-evident and convincing: Or if there were, could we advance a step beyond it, but by the use of those very faculties, of which we are supposed to be already diffident. The Cartesian doubt, therefore, were it ever possible to be attained by any human creature (as it plainly is not) would be entirely incurable; and no reasoning could ever bring us to a state of assurance and conviction upon any subject. (EHU 12.3 / 149–50)

But this result – as we have seen – gives no practical basis for scepticism. Certainly it raises a ground for theoretical concern, and highlights ‘the whimsical condition of mankind, who must act and reason and believe; though they are not able, by their most diligent enquiry, to satisfy themselves concerning the foundation of these operations’ 59

9780826443595_Ch03_Final_txt_print.indd 59

2/9/2012 9:38:49 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION for his notion of ‘reason’ and for the rational status of inductive inference. These issues are far from straightforward, partly because the argument appears three times in Hume’s works, with many differences between the three presentations – some of them highly significant – and clear evidence of a systematic development in his views. But for our purposes, it will be enough here just to highlight the most salient points.

(EHU 12.23 / 160). But unless we are in the grip of the rationalist prejudice that Hume rejects, we should not see this lack of theoretical satisfaction as sufficient reason to abandon our only respectable method of inference about the unobserved. That would be to take the sceptical considerations to a ridiculous (and anyway unachievable) extreme. Instead, the appropriate response is less dramatic but far more valuable: to recognize our ‘whimsical condition’ as a ground for modesty about the depth and extent of our powers, and to adopt a ‘mitigated scepticism’ which is correspondingly diffident and cautious (EHU 12.24 / 161–2), and which confines our attention to the subjects of common life, ‘avoiding distant and high enquiries’:

2.1 THE ARGUMENT OF THE TREATISE In the Treatise, the famous argument occurs within the context of Hume’s rather rambling search for the origin of the idea of necessary connexion, which he has previously (THN 1.3.2.11 / 77) identified as the key component of our idea of causation. Not having ‘any certain view or design’ on how to trace the impression(s) that could account for this crucial idea, he sets off to ‘beat about all the neighbouring fields’ in the hope that something will turn up (THN 1.3.2.12–13 / 77–8). His first such ‘field’ concerns the basis of the Causal Maxim ‘that whatever begins to exist, must have a cause of existence’ (THN 1.3.3.1 / 78), but after concluding that this Maxim cannot be ‘intuitively or demonstratively certain’,3 he quickly moves on to a related question, ‘Why we conclude, that such particular causes must necessarily have such particular effects, and why we form an inference from one to another?’ (THN 1.3.3.8–9 / 82). He soon narrows his focus onto what he considers the paradigm case of a causal inference, from a sensory impression of one ‘object’ (for example, we see a flame), to forming a belief – a lively idea – of its effect or cause (for example, we expect heat). He then analyses such an inference into its component parts: ‘First, The original impression. Secondly, The transition

While we cannot give a satisfactory reason, why we believe, after a thousand experiments, that a stone will fall, or fire burn; can we ever satisfy ourselves concerning any determination, which we may form, with regard to the origin of worlds, and the situation of nature, from, and to eternity? (EHU 12.25 / 162) This sentence is Hume’s last word on the question of inductive scepticism, and represents the conclusion of a coherent line of thought which can be traced from the beginning of Enquiry Section 12, his most clear and explicit – and repeatedly refined – treatment of scepticism as a whole. So far, then, we have a clear outline of his mature position.

2. HUME’S SCEPTICAL ARGUMENT The main aim of this chapter is to understand the logic and significance of Hume’s famous argument, and in particular its implications 60

9780826443595_Ch03_Final_txt_print.indd 60

2/9/2012 9:38:49 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION in the Treatise but disappears from his later writings.5 Hume has now established one of the most important results of his philosophy: ‘’Tis . . . by experience only, that we can infer the existence of one object from that of another.’ (THN 1.3.6.2 / 87). And he immediately goes on to explain that the kind of experience which prompts such a causal inference is repeated conjunctions of pairs of ‘objects . . . in a regular order of contiguity and succession’. Where we have repeatedly seen A closely followed by B, ‘we call the one cause and the other effect, and infer the existence of the one from that of the other’. Hume enthusiastically trumpets this relation of constant conjunction as the sought-for key to the crucial notion of necessary connexion, with a clear allusion back from THN 1.3.6.3 / 87 to 1.3.2.11 / 77,6 and he celebrates the progress of his rambling journey of discovery. Admittedly there is still some way to go, because mere repetition of conjunctions does not seem to generate ‘any new original idea, such as that of a necessary connexion’. But the line of investigation seems clear:

to the idea of the connected cause or effect. Thirdly, The nature and qualities of that idea.’ (THN 1.3.5.1 / 84). The remainder of Section 1.3.5 discusses the first component, then 1.3.6, entitled ‘Of the Inference from the Impression to the Idea’, comes to the second component, the causal inference itself.4 Hume’s first move, in discussing this paradigm causal inference, is to insist that it cannot be made a priori, simply from observation of the cause: There is no object, which implies the existence of any other if we consider these objects in themselves, and never look beyond the ideas which we form of them. Such an inference . . . wou’d imply the absolute contradiction and impossibility of conceiving any thing different. But as all distinct ideas are separable, ’tis evident there can be no impossibility of that kind. When we pass from a present impression to the idea of any object, we might possibly have separated the idea from the impression, and have substituted any other idea in its room. (THN 1.3.6.1 / 86–7)

having found, that after the discovery of the constant conjunction of any objects, we always draw an inference from one object to another, we shall now examine the nature of that inference . . . Perhaps ’twill appear in the end, that the necessary connexion depends on the inference, instead of the inference’s depending on the necessary connexion. (THN 1.3.6.3 / 88)

Here Hume is appealing to the principle that if an inference is to be a priori, there must be an absolute contradiction and impossibility of conceiving things as turning out differently: an a priori inference has to yield total certainty. He also seems to be taking for granted that such a contradiction in conception implies a contradiction in fact, which is closely related to his Conceivability Principle that whatever we conceive is possible (this makes a more explicit entrance shortly, at THN 1.3.6.5 / 89). Note also his appeal to what is commonly called his Separability Principle, that ‘all distinct ideas are separable’ (cf. THN 1.1.3.4 / 10, 1.1.7.3 / 18–19, 1.3.3.3 / 79–80), which plays a major role

This last sentence provides an elegant epitome of the link between Hume’s theories of induction and causation, anticipating the eventual outcome of his quest for the elusive impression of necessary connexion (which will come much later, at THN 1.3.14.20 / 164–5). For present purposes, however, we 61

9780826443595_Ch03_Final_txt_print.indd 61

2/9/2012 9:38:49 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION Conceivability Principle shows it cannot be: ‘We can at least conceive a change in the course of nature; which sufficiently proves, that such a change is not absolutely impossible. To form a clear idea of any thing, is an undeniable argument for its possibility, and is alone a refutation of any pretended demonstration against it.’ (THN 1.3.6.5 / 89). As for probable arguments (that is, arguments in which we draw conclusions – typically about things in the world of our everyday experience – with less than total certainty), these must be based on causal relations, because causation is ‘The only . . . relation of objects . . . on which we can found a just inference from one object to another’ (THN 1.3.6.7 / 89).9 But Hume has just argued that causal inference is ‘founded on the presumption of a resemblance betwixt those objects, of which we have had experience, and those, of which we have had none’ (an argument that he recapitulates at THN 1.3.6.6–7 / 89–90, echoing the discussion of THN 1.3.4.1–4 / 82–9).10 And since probable inference relies on causal relations, ‘’tis impossible this presumption [of the Uniformity Principle] can arise from probability’, on pain of circularity.11 So neither demonstrative nor probable arguments can provide any solid basis for the Uniformity Principle, and Hume quickly concludes that reason cannot be responsible for causal inference:12

can forget about that quest, and focus on the nature of inductive inference. Having established that causal inference ‘from the impression to the idea’ (e.g. from seeing A to expecting B) depends on experience, Hume goes on to pose the central question that his argument aims to answer, namely which mental faculty is responsible for the inference: ‘the next question is, Whether experience produces the idea by means of the understanding or imagination; whether we are determin’d by reason to make the transition, or by a certain association and relation of perceptions?’ (THN 1.3.6.4 / 88–9). If the faculty of reason were responsible, Hume says, this would have to be on the basis of an assumption of similarity between past and future, commonly called his Uniformity Principle: ‘If reason determin’d us, it wou’d proceed upon that principle, that instances, of which we have had no experience, must resemble those, of which we have had experience, and that the course of nature continues always uniformly the same.’ (THN 1.3.6.4 / 89). So the next stage is to see whether there is any argument by which reason could establish this principle, and if there is not, then Hume will conclude that reason cannot be the basis for our inductive inferences. Following the standard categorization deriving from John Locke,7 just two types of argument are potentially available, demonstrative and probable, and Hume now eliminates each in turn. First, demonstrative arguments proceed with absolute certainty based on self-evident (‘intuitive’) relationships between the ideas concerned; these sorts of argument are capable of yielding ‘knowledge’ in the strict sense, and are mostly confined to mathematics.8 But no such argument can possibly prove the Uniformity Principle, because that would mean the principle is absolutely guaranteed, which the

Thus not only our reason fails us in the discovery of the ultimate connexion of causes and effects, but even after experience has inform’d us of their constant conjunction, ’tis impossible for us to satisfy ourselves by our reason, why we shou’d extend that experience beyond those particular instances, which have fallen under our observation. We suppose, but are never able to prove, that there must be a resemblance betwixt those objects, 62

9780826443595_Ch03_Final_txt_print.indd 62

2/9/2012 9:38:49 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION of which we have had experience, and those which lie beyond the reach of our discovery. (THN 1.3.6.11 / 91–2)

afford us a reason for drawing a conclusion beyond it; and, that even after the observation of the frequent or constant conjunction of objects, we have no reason to draw any inference concerning any object beyond those of which we have had experience; . . . and this will throw them so loose from all common systems, that they will make no difficulty of receiving any, which may appear the most extraordinary. (THN 1.3.12.20 / 139)

Instead, such inference must derive from associative principles in the imagination (THN 1.3.6.12 / 92), and in particular, from a mechanism which Hume calls custom (e.g. THN 1.3.7.6 / 97, 1.3.8.10 / 102) or habit (e.g. THN 1.3.10.1 / 118). Experience of constant conjunction between A and B establishes an associative connexion between them, making our mind habitually move easily from the idea of one to the idea of the other. When we then see an A, the ‘force and vivacity’ of that sense impression is transferred through the associative link to our idea of B, enlivening it into a belief. Hume accordingly goes on to define a belief as ‘a lively idea related to or associated with a present impression’ (THN 1.3.7.5 / 96), and to expand on this theory of belief formation over the subsequent sections.

But this again is within a context where his aim is to develop his theory of belief, now focusing on inferences involving probability where the relevant past conjunctions are not constant. Books 1 and 2 of the Treatise were published at the end of January 1739, but well before the end of that year, Hume seems to have radically reassessed the significance of his philosophy. By then he had written his Abstract of the Treatise, which appeared in print in March 1740, and which devotes 8 paragraphs out of 35 (paragraphs 8 and 10–16) to the famous argument. From being a very small part of a much larger system, suddenly it becomes the prime focus of his philosophy, as it remained in the first Enquiry of 1748, which can indeed be seen as mainly constructed around the argument and its implications. The declared purpose of the argument in the Abstract is to understand ‘all reasonings concerning matter of fact’ (Abs. 8 / 649), rather than limiting discussion to the paradigm case of a causal inference – ‘the inference from the impression to the idea’ – which had been the topic of Treatise 1.3.6. But Hume then immediately states that all such factual reasonings (to coin a shorthand term) ‘are founded on the relation of cause and effect’, thus making clear that causal inference is still the focus. However, this initial

2.2 FROM THE TREATISE TO THE ABSTRACT Given the fame that it has subsequently enjoyed, Hume’s argument in Treatise 1.3.6 is surprisingly inconspicuous. It occurs within a detour (at THN 1.3.3.9 / 82) from a ramble through fields (THN 1.3.2.13 / 77–8); the core of it occupies only six fairly short paragraphs (1–2 / 86–7 and 4–7 / 88–90); and its primary role seems to be to identify custom as the ground of causal belief – as a component in Hume’s larger theory of belief – rather than to emphasize its own apparently sceptical conclusion. He does later remark on the striking nature of this conclusion:13 Let men be once fully perswaded of these two principles, that there is nothing in any object, consider’d in itself, which can 63

9780826443595_Ch03_Final_txt_print.indd 63

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION forthrightly elsewhere: ‘that to consider the matter a priori, any thing may produce any thing’ (THN 1.4.5.30 / 247, cf. 1.3.15.1 / 173, EHU 12.29 / 164). So experience is necessary to ground any causal inference (and hence any inference ‘concerning matter of fact’). And Hume goes on to explain that the type of experience relevant to his thought-experiment would be of ‘several instances’ (Abs. 12 / 651) in which Adam saw the collision of one ball into another followed by motion in the second ball. Such experience would condition him ‘to form a conclusion suitable to his past experience’, and thus to expect more of the same. ‘It follows, then, that all reasonings concerning cause and effect, are founded on experience, and that all reasonings from experience are founded on the supposition, that the course of nature will continue uniformly the same.’ (Abs. 13 / 651). So as in the Treatise, we reach Hume’s Uniformity Principle, and he now proceeds accordingly to consider what rational basis this principle could be given:

move is helpful in both emphasizing the generality of the argument and also streamlining it, avoiding the need for the recapitulation of his treatment of causal reasoning which had occupied THN 1.3.6.6–7 / 89–90. Now, in proving that all causal reasoning presupposes the Uniformity Principle, he will have proved at the same time that ‘all reasoning concerning matter of fact’ – and hence all probable reasoning – has such a dependence.14 To facilitate discussion, Hume introduces the simple example of one billiard ball striking another and causing it to move (Abs. 9–10 / 649–50). He then presents a vivid thoughtexperiment, imagining the first man Adam, newly created by God, and confronted with such an imminent collision: without experience, he would never be able to infer motion in the second ball from the motion and impulse of the first. It is not any thing that reason sees in the cause, which makes us infer the effect. Such an inference, were it possible, would amount to a demonstration, as being founded merely on the comparison of ideas. But no inference from cause to effect amounts to a demonstration. Of which there is this evident proof. The mind can always conceive any effect to follow from any cause, and indeed any event to follow upon another: whatever we conceive is possible, at least in a metaphysical sense: but wherever a demonstration takes place, the contrary is impossible, and implies a contradiction. There is no demonstration, therefore, for any conjunction of cause and effect. (Abs. 11 / 650–1)

’Tis evident, that Adam . . . would never have been able to demonstrate, that the course of nature must continue uniformly the same, and that the future must be conformable to the past. What is possible can never be demonstrated to be false; and ’tis possible the course of nature may change, since we can conceive such a change. (Abs. 14 / 651) As in the Treatise, we have an appeal to the Conceivability Principle to show that a change in the course of nature is possible, which in turn implies that uniformity cannot be demonstrated.

Compared to the equivalent passage in the Treatise (THN 1.3.6.1 / 86–7), this is clearer and more straightforward, proving by direct appeal to the Conceivability Principle a general lesson which he states even more

Nay, . . . [Adam] could not so much as prove by any probable arguments, that the 64

9780826443595_Ch03_Final_txt_print.indd 64

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION future must be conformable to the past. All probable arguments are built on the supposition, that there is this conformity betwixt the future and the past, and therefore can never prove it. This conformity is a matter of fact, and if it must be proved, will admit of no proof but from experience. But our experience in the past can be a proof of nothing for the future, but upon a supposition, that there is a resemblance betwixt them. This therefore is a point, which can admit of no proof at all, and which we take for granted without any proof. (Abs. 14 / 651–2)

how the world is, and so can be known (if at all) only through experience. Some matters of fact we learn directly by perception, and can later recall.16 But what of the rest? Hume sets himself to address this key question: ‘what is the nature of that evidence, which assures us of any real existence and matter of fact, beyond the present testimony of our senses, or the records of our memory’ (EHU 4.3 / 26)? On what basis do we infer from what we perceive and remember, to conclusions about further, unobserved, matters of fact? Hume calls such inferences ‘reasonings concerning matter of fact’ (EHU 4.4 / 26), a term we saw introduced just once in the Abstract but which now becomes his standard way of referring to what he had previously called ‘probable arguments’. The reason for this terminological adjustment seems to be to avoid the infelicity of calling such inferences merely ‘probable’ even when they are based on vast and totally uniform past experience that yields complete ‘moral certainty’ (that is, practical assurance). In a footnote to the heading of Section 6, Hume will accordingly draw a distinction – within the class of ‘reasonings concerning matter of fact’ – between probabilities and proofs, the latter being ‘such arguments from experience as leave no room for doubt or opposition’, as when we conclude that ‘all men must die, or that the sun will rise to-morrow’.17 In Enquiry 4, the famous argument now proceeds much as it had in the Abstract, albeit greatly filled out. The appendix to this chapter lays out a structure diagram involving 20 stages,18 with the stages numbered according to the logic of the argument. The same numbers will be followed here, within square brackets, to enable easy cross-referencing. First, we learn that [2] ‘All reasonings concerning matter of fact seem to be founded on the relation of Cause and Effect’ (EHU 4.4 / 26),

Here the logical circularity of attempting to give a probable argument for the Uniformity Principle is more explicitly spelled out than in the Treatise. With both demonstrative and probable argument eliminated, Hume briskly concludes that ‘We are determined by custom alone to suppose the future conformable to the past. . . . ’Tis not, therefore, reason, which is the guide of life, but custom.’ (Abs. 15–16 / 652).15 2.3 THE ARGUMENT OF THE ENQUIRY In the Enquiry concerning Human Understanding of 1748, the famous negative argument occupies virtually all of Section 4, with the positive account in terms of custom appearing in Section 5. Compared with the versions in the Treatise and Abstract, the argument is clarified and greatly expanded, leaving little doubt that Hume considers this his definitive presentation. Section 4 starts with an important distinction now commonly known as ‘Hume’s Fork’, between relations of ideas – that is, propositions (notably from mathematics) that can be known to be true a priori, just by examining and reasoning with the ideas concerned – and matters of fact – that is, propositions whose truth or falsehood depends on 65

9780826443595_Ch03_Final_txt_print.indd 65

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION discovered in the cause, and the first invention or conception of it, a priori, must be entirely arbitrary. And even after it is suggested, the conjunction of it with the cause must appear equally arbitrary; since there are always many other effects, which, to reason, must seem fully as consistent and natural. (EHU 4.11 / 30)

since [1] ‘By means of that relation alone we can go beyond the evidence of our memory and senses.’ As in the Abstract, starting in this way has the virtue of streamlining the argument that follows, so that conclusions to be drawn about causal reasoning will automatically apply to the entire class of factual reasoning. The first of these conclusions, as before, is that [5] all knowledge of causal relations must be founded on experience: ‘the knowledge of this relation [i.e. causation] is not, in any instance, attained by reasonings a priori; but arises entirely from experience, when we find, that any particular objects are constantly conjoined with each other.’ (EHU 4.6 / 27). Again we get a thought-experiment involving Adam, but this time with water and fire, illustrating the general truth that [3] ‘No object ever discovers [i.e. reveals], by the qualities which appear to the senses, either the causes which produced it, or the effects which will arise from it’. This is relatively easy to see when the phenomena are untypical or unfamiliar, such as the unexpected adhesion between smooth slabs of marble, the explosion of gunpowder, or the powers of a (magnetic) lodestone, where we have no temptation to imagine that we could have predicted these effects in advance (EHU 4.7 / 28). But with commonplace occurrences, such as the impact of billiard balls (EHU 4.8 / 28–9), we might suppose that the effect was foreseeable a priori. To prove that this is an illusion, Hume asks us to imagine how we could possibly proceed to make such an a priori inference, arguing that we could not, on the grounds that the effect is a quite distinct event from the cause (EHU 4.9 / 29), while many different possible effects are equally conceivable (EHU 4.10 / 29–30). Summing up [4]:

Note the strong emphasis on arbitrariness, making clear that it is not just the conceivability – or mere theoretical possibility – of alternative outcomes which makes any a priori inference from cause to effect impossible; it is the fact that from an a priori point of view, there is nothing to suggest one outcome over another.19 If causal relations cannot be known a priori, then factual inference cannot be a priori either (given [2] that factual inference is founded on causation). [6] ‘In vain, therefore, should we pretend to determine any single event . . . without the assistance of observation and experience.’ Hume now brings Part 1 of Section 4 to a close, with two very important corollaries for his philosophy of science. The first is that since we cannot aspire to a priori insight into why things work as they do, the appropriate ambition for science is instead to aim more modestly for systematization of those cause and effect relationships that experience reveals: ‘to reduce the principles, productive of natural phænomena, to a greater simplicity, and to resolve the many particular effects into a few general causes, by means of reasonings from analogy, experience, and observation.’ (EHU 4.12 / 30). Then follows Hume’s most explicit account of applied mathematics (which he calls ‘mixed mathematics’), emphasizing that although mathematical relationships are a priori, the laws through which they are applied to the world – his example is the Newtonian law of

every effect is a distinct event from its cause. It could not, therefore, be 66

9780826443595_Ch03_Final_txt_print.indd 66

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION This passage seems to be saying that [7] when we draw conclusions from past experience, we presuppose a resemblance between the observed and the unobserved, extrapolating from one to the other.22 Later, when apparently referring back to this passage, Hume confirms such a reading: ‘We have said, . . . that all our experimental conclusions proceed upon the supposition, that the future will be conformable to the past.’ (EHU 4.19 / 35). So his ‘main question’ at EHU 4.16 / 34 concerns, in effect, the foundation of the Uniformity Principle.23 He repeats (cf. EHU 4.6 / 27) that [3] ‘there is no known connexion between the sensible qualities and the secret powers’ of any object, and infers from this that [9] ‘the mind is not led to form such a conclusion concerning their constant and regular conjunction, by any thing which it knows of their nature’ (EHU 4.16 / 33). So the Uniformity Principle cannot be established on the basis of anything that we learn directly through sense perception, in which case [10] any foundation for it will have to draw on past experience, which for the sake of the argument can here be taken as infallible: ‘As to past Experience, it can be allowed to give direct and certain information of those precise objects only, and that precise period of time, which fell under its cognizance . . . . (EHU 4.16 / 33). The ‘main question’ is then urged: how to justify the step from past experience to the assumption of future resemblance?

conservation of momentum – remain unambiguously a posteriori: ‘the discovery of the law itself is owing merely to experience, and all the abstract reasonings in the world could never lead us one step towards the knowledge of it’ (EHU 4.13 / 31).20 Part 2 starts by summarizing Hume’s results so far, and anticipating his eventual conclusion [20]: When it is asked, What is the nature of all our reasonings concerning matter of fact? the proper answer seems to be, that they are founded on the relation of cause and effect. When again it is asked, What is the foundation of all our reasonings and conclusions concerning that relation? it may be replied in one word, Experience. But if we still carry on our sifting humour, and ask, What is the foundation of all conclusions from experience? this implies a new question . . . I shall content myself, in this section, with an easy task, and shall pretend [i.e. aspire] only to give a negative answer to the question here proposed. I say then, that, even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding. (EHU 4.14–15 / 32) Having established that experience is required for any factual inference, Hume goes on to explain how experience plays that role: we always presume, when we see like sensible qualities, that they have like secret powers,21 and expect, that effects, similar to those which we have experienced, will follow from them. . . . But why [past] experience should be extended to future times, and to other objects, which for aught we know, may be only in appearance similar; this is the main question . . . (EHU 4.16 / 33–4; emphasis added)

These two propositions are far from being the same, I have found that such an object has always been attended with such an effect, and I foresee, that other objects, which are, in appearance, similar, will be attended with similar effects. I shall allow, if you please, that the one proposition may justly be inferred from the other: I know in fact, that it always is 67

9780826443595_Ch03_Final_txt_print.indd 67

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION can never be proved false by any demonstrative argument or abstract reasoning a priori. (EHU 4.18 / 35)

inferred. But if you insist, that the inference is made by a chain of reasoning, I desire you to produce that reasoning. The connexion between these propositions is not intuitive. There is required a medium, which may enable the mind to draw such an inference, if indeed it be drawn by reasoning and argument. (EHU 4.16 / 34)

As in the Treatise and Abstract, Hume appeals to the Conceivability Principle, though slightly differently: here he expresses it as the principle that what is conceivable implies no contradiction, rather than saying that what is conceivable is possible.26 Moving on now to probability:

So because [11] the inference from past experience to future resemblance is not intuitive (i.e. not immediately self-evident), [12] there must be some medium, some ‘connecting proposition or intermediate step’ (EHU 4.17 / 34) if indeed the inference is ‘drawn by reasoning and argument’.24 The long paragraph that we have just been discussing (EHU 4.16 / 32–4) includes steps that have no parallel in the Treatise and Abstract, where, as we saw, Hume simply takes for granted that if the Uniformity Principle is to be rationally well founded, then this must be on the basis of some chain of reasoning, either demonstrative or probable. Here in the Enquiry, he explicitly rules out both sense experience and intuition as sources of foundation for the Uniformity Principle, and only then comes to consider demonstration and probability, which are in turn dismissed in the familiar way, but again with the structure of the argument made somewhat more explicit:

[16] If we be, therefore, engaged by arguments to put trust in past experience, and make it the standard of our future judgment, these arguments must be probable only, or such as regard matter of fact and real existence, . . . But . . . there is no argument of this kind, . . . We have said, that [2] all arguments concerning existence are founded on the relation of cause and effect; that [5] our knowledge of that relation is derived entirely from experience; and that [7] all our experimental conclusions proceed upon the supposition, that the future will be conformable to the past. [17] To endeavour, therefore, the proof of this last supposition by probable arguments, or arguments regarding existence, must be evidently going in a circle, and taking that for granted, which is the very point in question. (EHU 4.19 / 35–6)

[13] All reasonings may be divided into two kinds, namely demonstrative reasoning, or that concerning relations of ideas, and moral reasoning, or that concerning matter of fact and existence.25 [15] That there are no demonstrative arguments in the case, seems evident; since [14] it implies no contradiction, that the course of nature may change, . . . Now whatever is intelligible, and can be distinctly conceived, implies no contradiction, and

Note in passing how Hume just assumes here some obvious inferences, linking [2] with [5] to deduce that [6] all factual inferences (‘probable arguments’, ‘arguments concerning existence’) are founded on experience, and then combining this with [7] to deduce in turn that [8] all factual inferences ‘proceed upon the supposition’ of the Uniformity Principle.27 He also now leaves the reader to piece together the final 68

9780826443595_Ch03_Final_txt_print.indd 68

2/9/2012 9:38:50 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION (A) The argument concerns all inferences to matters of fact that we have not observed: what the Enquiry calls ‘reasonings concerning matter of fact’ (here factual inferences for short). Although the Treatise version starts with a narrower focus on causal inference ‘from the impression to the idea’, it later requires the lemma that all factual inferences are based on causal relations (stated at THN 1.3.6.7 / 89). So the argument is improved both structurally and philosophically by starting with all factual inferences, as in the Abstract and the Enquiry, and then deriving this lemma as its first main stage (Abs. 8 / 649; EHU 4.4 / 26–7). (B) Hume next argues that causal relations cannot be known a priori, and hence are discoverable only through experience (THN 1.3.6.1 / 86–7, Abs. 9–11 / 649–51; EHU 4.6–11 / 27–30). This is a major principle of his philosophy, wielded significantly elsewhere (e.g. THN 1.3.15.1 / 173, 1.4.5.30 / 247–8; EHU 12.29 / 164). (C) From this principle, together with the lemma from (A), Hume concludes that all factual inferences are founded on experience, the relevant experience being of those constant conjunctions through which we discover causal relationships (THN 1.3.6.2 / 87, Abs. 12 / 651; EHU 4.16 / 33). (D) Factual inferences thus involve extrapolation from observed to unobserved, based on an assumption of resemblance between the two. Initially in the Treatise, Hume seems to suggest that such an assumption of resemblance – commonly called his Uniformity Principle (UP) – would be necessarily implicated only if reason were responsible for the inference (THN 1.3.6.4 / 88–9). But his settled view, expressed in all three works (see note 10 above), is that UP is presupposed by all factual

stages of his argument.28 First, that since the Uniformity Principle cannot be established by either demonstrative or factual inference, it follows that [18] there is no good argument for the Uniformity Principle. Secondly, that therefore (given [12]),29 it follows that [19] the Uniformity Principle cannot be founded on reason, and finally, that since [8] all factual inferences are founded on the Uniformity Principle, it follows that [20] no factual inference (i.e. no ‘reasoning concerning matter of fact and existence’) is founded on reason. Hume had anticipated this conclusion at EHU 4.15, quoted earlier:30 ‘I say then, that, even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding.’ (EHU 4.15 / 32). Also in the following section – most of which is devoted to sketching his theory of belief as based on ‘Custom or Habit’ (EHU 5.5 / 43) – Hume refers back to this argument and states its conclusion explicitly, once purely negatively and once alluding to his positive theory: ‘we . . . conclude . . . in the foregoing section, that, in all reasonings from experience, there is a step taken by the mind, which is not supported by any argument or process of the understanding; . . .’ (EHU 5.2 / 41); ‘All belief of matter of fact or real existence . . . [is due merely to] . . . a species of natural instincts, which no reasoning or process of the thought and understanding is able, either to produce, or to prevent.’ (EHU 5.8 / 46–7). 2.4 THE ESSENTIAL CORE OF HUME’S SCEPTICAL ARGUMENT

We can now distil the essence of Hume’s argument from these three different presentations, into eight main stages: 69

9780826443595_Ch03_Final_txt_print.indd 69

2/9/2012 9:38:51 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION inferences,31 simply in virtue of their taking for granted a resemblance between observed and unobserved. (E) Hume now proceeds to investigate critically the basis of UP itself. In the Treatise (THN 1.3.6.4 / 88–9) and Abstract (Abs. 14 / 651–2), he appears to assume immediately that any foundation in reason would have to derive from some demonstrative (i.e. deductive) or probable (i.e. factual) inference. In the Enquiry, however – which hugely expands this part of the argument from the cursory treatment in the earlier works – he considers demonstrative and factual inference only after first (EHU 4.16 / 32–4) explicitly ruling out any foundation in sensory awareness of objects’ powers, or in immediate intuition (i.e. self-evidence).32 (F) Any demonstrative argument for UP is ruled out because a change in the course of nature is clearly conceivable and therefore possible (THN 1.3.6.5 / 89; Abs. 14 / 651–2, EHU 4.18 / 35). Any factual argument for UP is ruled out because, as already established at (D), such arguments inevitably presuppose UP, and hence any purported factual inference to UP would be viciously circular (THN 1.3.6.7 / 89–90; Abs. 14 / 651–2; EHU 4.19 / 35–6). (G) The upshot of this critical investigation is that UP has no satisfactory foundation in reason, though Hume expresses this in various ways:

This [resemblance between past and future] is a point, which can admit of no proof at all, and which we take for granted without proof. (Abs. 14 / 652) it is not reasoning which engages us to suppose the past resembling the future, and to expect similar effects from causes, which are, to appearance, similar. (EHU 4.23 / 39) (H) Since UP is presupposed by all factual inferences (D), and UP has no foundation in reason (G), Hume finally concludes that factual inference itself has no foundation in reason. Again he expresses this conclusion in various ways (and note here the narrower focus of the Treatise on causal inference ‘from the impression to the idea’, as pointed out at (A) above): When the mind . . . passes from the idea or impression of one object to the idea or belief of another, it is not determin’d by reason (THN 1.3.6.12 / 92) ’Tis not, therefore, reason, which is the guide of life, but custom. That alone determines the mind . . . to suppose the future conformable to the past. However easy this step may seem, reason would never, to all eternity, be able to make it. (Abs. 16 / 652) I say then, that, . . . our conclusions from . . . experience are not founded on reasoning, or any process of the understanding. (EHU 4.15 / 32)

’tis impossible to satisfy ourselves by our reason, why we shou’d extend [our] experience beyond those particular instances, which have fallen under our observation. We suppose, but are never able to prove, that there must be a resemblance betwixt those objects, of which we have had experience, and those which lie beyond the reach of our discovery. (THN 1.3.6.11 / 91–2)

in all reasonings from experience, there is a step taken by the mind, which is not supported by any argument or process of the understanding (EHU 5.2 / 41) Note also that two of these quotations – from Abs. 16 / 652 and EHU 5.2 / 41 – could just as appropriately have been cited as illustrations of (G), because both refer to that ‘step’ 70

9780826443595_Ch03_Final_txt_print.indd 70

2/9/2012 9:38:51 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION glosses the conclusion of the argument in apparently very negative terms, as showing that ‘we have no reason’ to draw any factual inference (THN 1.3.12.20 / 139), and that ‘we cannot give a satisfactory reason, why we believe, after a thousand experiments, that a stone will fall, or fire burn’ (EHU 12.25 / 162). In this light, it seems entirely appropriate that he should entitle Enquiry Section 4 ‘Sceptical Doubts concerning the Operations of the Understanding’, and describe it as appearing to give the sceptic ‘ample matter of triumph’ (EHU 12.22 / 159). As discussed earlier, however, the issue of Hume’s inductive ‘scepticism’ is not so straightforward, and it is far from clear that he sees the acknowledged incapacity of reason to ‘prove’ or ‘support’ the Uniformity Principle as any sort of genuine problem. Certainly he does not infer from it (either in the Treatise, the Abstract or the Enquiry) that induction is unreasonable in any pragmatic sense. And indeed the line of thought sketched in section 1 above, drawing on Section 12 of the Enquiry, somewhat suggests that he considers it inevitable that our most basic principles of inference – precisely because they are so basic – will lack any ultimate justification beyond their fundamental place in our mental economy. That being so, the central upshot of Hume’s argument might be simply to identify the Uniformity Principle as a basic principle of this kind, and the sceptical flavour of his reasoning – in demonstrating reason’s incapacity to prove UP – need not carry over at all into the theory of human inference that he draws from it. Nevertheless, the sceptical flavour of the famous argument itself would remain, in denying UP a source of rational support that more optimistic philosophers might have expected it to enjoy. And although the argument also delivers the important positive principle that

which is precisely the presupposition of the Uniformity Principle. Since factual inference operates by extrapolation from past to future, Hume takes it to be obvious that the foundation of such inference must be the same as the foundation of the principle of extrapolation. Hence he does not consistently distinguish between (G) and (H), making the last stages of his argument less explicit than one might wish (cf. the end of section 2.3 above).

3. THE NATURE OF HUME’S SCEPTICAL CONCLUSION Hume usually expresses the conclusion of his famous argument in a way that seems to imply some incapacity on the part of human reason. The Uniformity Principle is something that we ‘are never able to prove’ (THN 1.3.6.11 / 92), and which indeed ‘can admit of no proof at all’ (Abs. 14 / 652). Because of this, ‘’tis impossible to satisfy ourselves by our reason’ (THN 1.3.6.11 / 91) concerning the inferential step from past to future, a step which ‘reason would never, to all eternity, be able to make’ (Abs. 16 / 652) and ‘which is not supported by any argument or process of the understanding’ (EHU 5.2 / 41). Hume also frequently uses similar terms within the argument itself, when saying that various would-be proofs of UP are impossible, refutable, circular or lack any ‘just foundation’ (THN 1.3.6.5 / 89, 1.3.6.7 / 89–90, 1.3.6.10 / 91; EHU 4.18 / 35, 19 / 35–6, 21 / 37–8), denying that human knowledge ‘can afford . . . an argument’ that ‘supports the understanding’ (EHU 4.17 / 34) in reasoning from past to future, and consequently denying that our factual inferences ‘are built on solid reasoning’ (THN 1.3.6.8 / 90). In both the Treatise (see section 2.2 above) and Enquiry (see section 1), he later 71

9780826443595_Ch03_Final_txt_print.indd 71

2/9/2012 9:38:51 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION reject some non-Humean notion of reason; indeed this style of interpretation became extremely popular in the 1980s. Before then, the general image of Hume was of a highly destructive sceptic, intending through his famous argument to maintain that inductive arguments lack all rational justification. Barry Stroud wittily expressed what he took to be Hume’s conclusion: ‘As far as the competition for degrees of reasonableness is concerned, all possible beliefs about the unobserved are tied for last place.’36 Some of these extreme sceptical interpretations – most influentially those of Antony Flew37 and David Stove38 – took Hume to be starting from the assumption of deductivism, that an inference is rationally justified only if it is logically guaranteed.39 But deductivism sits very uneasily with the (fallible but reasonable) empirical judgements that abound within Hume’s contributions to ‘the science of human nature’, for example his discussions of the passions, his Essays on politics and economics, and his various pieces on religion. The ‘wise man’ of Enquiry 10.4 / 110, who ‘proportions his belief to the [empirical] evidence’, clearly cannot be a deductivist; hence Flew, in discussing ‘Of Miracles’, was forced to accuse Hume of ‘flagrant and embarrassing’ inconsistency.40 Even more flagrantly inconsistent, from this perspective, were the passages in which Hume, after his famous argument had apparently denied causal, factual inference a place within the realm of ‘reason’, then quite explicitly treated it as one of reason’s central operations, for example:41

the Uniformity Principle is presupposed by all factual inference (D), even in the Enquiry we have to wait until the final section to see this wielded as part of an effective theoretical defence against the ‘Pyrrhonian’ sceptic.33 3.1 DEBATES ABOUT HUMEAN ‘REASON’ What we get much sooner, of course, and in all three works, is Hume’s positive account of how our inductive inferences operate through custom or habit – what he calls in the title of Enquiry Section 5 his ‘sceptical solution’ to the earlier ‘sceptical doubts’. But as David Owen observes, it seems odd to suppose that a psychological account of how belief functions could in any way ‘solve’ genuine epistemological doubts; Owen accordingly suggests that the famous argument is itself best understood as exclusively concerned with psychological mechanisms, and as having nothing to do with ‘the warrant of probable reasoning or the justification of belief’.34 The argument’s conclusion, that factual inference is not founded on reason, may initially seem unambiguously normative, but Owen interprets ‘reason’ here as signifying what he takes to be a Lockean conception of the mechanism by which human reasoning operates, namely, through stepwise inference via intermediate ideas. Thus he is able to read Hume’s conclusion that factual inference ‘is not determin’d by reason’ as purely descriptive: as denying that we actually draw factual inferences in the stepwise, mediated manner that Locke supposes.35 The famous argument accordingly serves to reject this Lockean conception of probable reasoning in favour of the more immediate and instinctive Humean model, thus providing a contribution to empirical psychology, rather than an exercise in sceptical epistemology. Owen is by no means the first to read Hume’s famous argument as designed to

with regard to reason . . . The only conclusion we can draw from the existence of one thing to that of another, is by means of the relation of cause and effect (THN 1.4.2.47 / 212) 72

9780826443595_Ch03_Final_txt_print.indd 72

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION reason, in a strict and philosophical sense, can have an influence on our conduct . . . by informing us of the existence of something which is a proper object of [a passion]; or when it discovers the connexion of causes and effects, so as to afford us means of exerting any passion. (THN 3.1.1.12 / 459)

proposed alternative was to see the argument as presupposing a perceptual model of reason, according to which we draw inferences through the perception of evidential connexions. This had the virtue of identifying a plausible (and substantial) target of Hume’s argument, namely, Locke’s view – complacently assumed and stated in his Essay concerning Human Understanding47 but never worked out in any detail – that probable reasoning is founded on the perception of probable connexions. Against this, Hume’s rival model of probable inference based on custom – and introduced immediately after his famous argument had refuted the alternative – stood out as a radical (and highly sceptical) departure. Moreover the ubiquitous hold of the traditional perceptual view of reason, which goes back to the ancients and was shared by Hume’s contemporaries, could help to make sense of his own apparent assessment of the famous argument as having a significant sceptical impact. A mere denial that induction has deductive force, or yields total certainty, would hardly be worthy of notice in the wake of Locke.48 But denying that induction is founded on perception of any good reason whatever would have vastly more sceptical significance. Garrett’s approach was quite different, and in context more radical. He insisted – against the prevailing orthodoxy – that Hume employs but a single sense of ‘reason’, taking this to be Hume’s name for ‘the general faculty of making inferences or producing arguments – just as it was for Locke’.49 Leaving aside the (debatable) attribution to Locke,50 the obvious advantage of this interpretation was precisely its lack of any need to posit an ambiguity in ‘reason’:

But such gross inconsistency was hard to credit, and given Hume’s evident commitment to inductive moral science (and to its reasonableness in comparison with ‘superstition’), it seemed to most later scholars far more plausible to interpret his famous argument not as genuinely sceptical, but instead as a way of rejecting or undermining the deductivist notion of reason on which it was thought to be based (by revealing its complete incapacity to underwrite any factual inference). From this it would follow that Hume must use the term ‘reason’ in at least two distinct senses: one narrowly deductivist or ‘rationalistic’ notion within the famous argument, and a broader, more ‘naturalistic’ notion elsewhere, such as in his discussions of the passions and morality. This anti-deductivist style of interpretation was pioneered in 1975 by Tom Beauchamp and Thomas Mappes,42 whose work was quickly followed by numerous variations on the theme.43 However it lost favour after I and Don Garrett (independently, in 1995 and 1997)44 pointed out the implausibility, under careful analysis, of reading Hume’s argument as employing a deductivist notion of reason. Deductivism proves very hard to square with the argument’s logic,45 and it also seems strange – if Hume’s purpose is to wield the argument in order to reject the notion of reason that it employs – that he should then continue to assert its sceptical conclusion in the apparently sincere terms we saw earlier.46 My own

Hume . . . [is] making a specific claim, within cognitive psychology, about the relation between our tendency to make 73

9780826443595_Ch03_Final_txt_print.indd 73

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION is . . . that [inductive inferences] are reasonings which are not themselves caused by any piece of reasoning (including, of course, themselves). Inductive inferences require that we bridge a gap between observation and prediction, and for someone not already disposed by nature to bridge that gap, no argument for doing so would be persuasive. Hence, . . .53

inductive inferences and our inferential/ argumentative faculty: he is arguing that we do not adopt induction on the basis of recognizing an argument for its reliability, for the utterly sufficient reason that there is no argument (‘reasoning’ or ‘process of the understanding’) that could have this effect. . . . this does not mean that inductive inferences are not themselves instances of argumentation or reasoning . . . His point is rather that they are reasonings that are not themselves produced by any piece of higher level reasoning: there is no argument that could lead us to accept the conclusion that inductive reasonings will be reliable if we did not already accept that conclusion in practice. Hence, in just this sense, they are a class of ‘reasonings’ (inferences or arguments) that ‘reason’ (the faculty of making inferences or giving arguments) does not itself ‘determine’ (cause) us to make.51

One surprising effect of this change was to bring Garrett’s interpretation rather close to Owen’s, because his detailed analysis of how mediated inferences operate made them ipso facto inferences that are ‘determin’d by reason’. On this account, a demonstrative inference from A to D, mediated by the intuitive connexions of A to B, B to C, and C to D, will include as part of its processing the intermediate inference connecting B to D.54 This makes the latter inference, according to Garrett, a cause of the overall inference from A to D; hence that overall inference is ‘determin’d by reason’ in the sense of being caused by another inference. Thus Garrett agrees with Owen that Hume’s conclusion involves a denial that induction proceeds by stepwise ratiocination. All this may seem somewhat artificial, and increasingly distant from anything to be found in Hume’s text, but it has the nice feature of accommodating a genuinely Humean point – that probable inference is characteristically immediate and instinctive rather than mediated and reflective – within a framework which, unlike Owen’s, avoids any need to treat the notion of ‘reason’ that is operative in the famous argument as one that Hume rejects. Garrett has consistently urged this last point against rival interpretations: that Hume’s famous argument gives little internal clue that he is employing some special notion of reason which he aims to reject. And although both Owen and I sought to mitigate the impact of

Like Owen, Garrett saw Hume’s argument as an exercise in descriptive psychology rather than normative epistemology, delivering a result about the causation of inductive inference rather than its ‘evidentiary value’. However, this sat uneasily with Garrett’s emphasis (as in the quotation above) on the recognition of higher-level arguments for the reliability of induction, and under challenge, he modified his original reading of Hume’s conclusion to make it more general. Here are the relevant parts of the passage above, edited to reflect the adjusted reading:52 Hume . . . [is] making a specific claim, . . . about the underlying causal mechanism that gives rise to inductive inferences: namely, that it is not itself dependent on any reasoning or inference. . . . His point 74

9780826443595_Ch03_Final_txt_print.indd 74

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION which means ‘Having the power of discovering truth immediately without ratiocination’. All this seems to fit with Hume’s own usage: he refers to ‘deductions’ and ‘ratiocination’ in contexts where stepwise argument is clearly intended (e.g. THN 1.3.14.2 / 156, EHU 5.22 / 55, EPM 1.4 / 170; EHU 4.23 / 39, 12.17 / 155), and he is happy to refer to ‘arguments’, ‘inference’ and ‘proof’ that are ‘intuitive’, and hence do not proceed in a stepwise fashion (THN 1.3.14.35 / 173, 2.3.2.2 / 408; EHU 4.21 / 37, 8.22n18 / 94n, LDH 1.187, 91).58 Overall, therefore, the language in which Hume expresses his famous conclusion is no argument (sic.) in favour of Garrett’s interpretation. Perhaps the most serious problem for Garrett’s interpretation, as for Owen’s, has been in making sense of the logic of Hume’s famous argument. For as we have seen, that argument does not in fact put much emphasis on a general absence of stepwise processing or ratiocination within inductive inference.59 Instead, it focuses on the very specific step of extrapolation from observed to unobserved – that is, the supposition of the Uniformity Principle (UP) – and then it attacks in turn the props on which that principle ‘may be suppos’d to be founded’, showing that none of them can ‘afford any just conclusion of this nature’ (THN 1.3.6.4 / 90). This move makes perfect sense on an epistemological interpretation of the argument, because if any essential evidential step in an inductive inference lacks a ‘just foundation’, then the inference as a whole will, apparently, be undermined.60 On the Garrett/Owen style of psychological interpretation, however, the move looks almost irrelevant – even if it sufficed to show that UP is not itself reached through mediated ratiocination, that would not exclude such ratiocination from playing some other role within inductive inference. This objection can

this criticism on our interpretations (by stressing that we saw Hume’s response to the argument as changing not the scope of his notion of reason, but rather its presumed method of operation),55 the lack of any obvious and deliberate ambiguity or equivocation on Hume’s part has remained by far the strongest weapon in Garrett’s armoury. Garrett’s interpretation has also seemed attractive for a more specious reason, namely, the extent to which Hume expresses his conclusion in terms of the impossibility of founding induction on ‘reasoning’ (e.g. THN 1.3.6.8 / 90; EHU 4.15 / 32, 4.16 / 34, 4.23 / 39), ‘proof’ (e.g. THN 1.3.6.11 / 92; Abs. 14 / 651–2), or ‘argument’ (e.g. Abs. 15 / 652; EHU 4.16–17 / 34–5, 4.21–3 / 38–9, 5.2 / 41). Today we read these terms as signifying complex inference involving intermediate steps, but in Hume’s day they were understood rather differently. Johnson’s dictionary of 1756 tells us that ‘reasoning’ is derived from ‘reason’, and defines it simply as ‘argument’.56 The first sense of ‘argument’ is given as ‘A reason alleged for or against any thing’, and Johnson implicitly confirms that he takes this as its primary sense in specifying – as one of the non-discursive senses of ‘reason’ – ‘Argument; ground of persuasion; motive’. Likewise the first sense of ‘proof’ is given as ‘Evidence; testimony; convincing token’, supplemented in later editions by the clauses ‘convincing argument; means of conviction’. The words that Johnson favours for stepwise inference are ‘deduction’ and ‘ratiocination’,57 as in the specification of the two discursive senses of ‘reason’: ‘The power by which man deduces one proposition from another, or proceeds from premises to consequences’, and ‘Ratiocination; discursive power’. These both contrast with ‘intuition’, which is ‘Knowledge not obtained by deduction of reason’, and ‘intuitive’, 75

9780826443595_Ch03_Final_txt_print.indd 75

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION has shown that UP has no ‘just’ foundation in demonstrative or factual reasoning.65 But why should this be thought to exhaust the possibilities of relevant reasoning? Hume quite often refers to ‘arguments’ or ‘reasoning’ that he considers ‘absurd’, ‘fallacious’, or ‘sophistical’ (e.g. THN 1.2.4.11 / 43, 1.4.5.30 / 247; DNR 9.189–92); on a psychological interpretation of the famous argument, these should be as relevant to his theory as the ‘just’ reasonings that he is able to rule out.66 Suppose, for example, it were suggested that induction can be founded on the principle that every change must have a cause, and hence the ultimate causal laws must be consistent over time. This would bring into play the attempted demonstrations of the Causal Maxim that Hume refutes at THN 1.3.3.4–8 / 80–2: even if fallacious, they could still be contenders as psychological explanations of our inductive assumptions. In short, Hume has done nothing to refute the hypothesis that UP may be believed on the basis of an invalid demonstrative argument; hence on the interpretations of Owen and Garrett, his famous argument is hopelessly incomplete.67 But things are even worse than this, for yet another strategy that remains open on their readings was actually used by Hume’s friend and correspondent Richard Price in the first chapter of his Review of the Principal Questions in Morals. Price claims that the Causal Maxim is known intuitively, ‘nothing being more clearly absurd and contradictory, than the notion of a change without a changer’.68 Then, in a footnote a few pages later, he explains how this can provide a basis for inferring future events from past regularity:

be sharpened by posing a dilemma over what role UP itself is supposed to play here.61 If Hume is saying that UP functions as an intermediate step in inductive inference, then it looks as though he thinks inductive inference does involve stepwise ratiocination (via UP itself), in which case he is contradicting the very conclusion that Owen takes him to be drawing from the argument. If, on the other hand, Hume is denying that UP can play any psychological role within inductive inference (on the basis that it has no ‘just foundation’), then it is unclear why he should take this to imply anything further about the actual psychological mechanism of inductive inference. The only apparently plausible answer here is to see Hume as placing a conditional constraint on how stepwise ‘reason’ could work: ‘If reason determin’d us, it wou’d proceed upon’ UP (THN 1.3.6.4 / 89; emphasis added). But this conditional statement appears only in the Treatise presentation, and even there, it is followed three paragraphs later by the unconditional statement that ‘probability is founded on the presumption of’ UP (THN 1.3.6.7 / 90; emphasis added).62 Owen largely builds his interpretation around the conditional,63 but it is straining credibility to rely so heavily on one statement in the Treatise, when we have seen so much evidence in sections 2.1–3 above that the versions in the Abstract and (especially) the Enquiry are more carefully crafted. On a psychological interpretation, moreover, Hume should not be so confident that mediated ratiocination for a factual conclusion can proceed only via UP.64 At best, he can claim that any rationally sufficient reasoning for such a conclusion must involve UP. And likewise, his argument seems completely inadequate to show that UP itself could not be believed on the basis of some mediated ratiocination. At best, again, he

The conviction produced by experience is built on the same principle, with that 76

9780826443595_Ch03_Final_txt_print.indd 76

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION we do than with why those beliefs are unjustified.71

which assures us, that there must be a cause of every event . . . Because we see intuitively, that there being some reason or cause of this constancy of event, it must be derived from causes regularly and constantly operating . . . And the more frequently and uninterruptedly we knew this had happened, the stronger would be our expectation of its happening again, . . . 69

But when articulating Hume’s conclusion, Owen apparently moves towards a claim about the functioning of all individual inductive inferences, and strongly contrasts this with the alternative view as ascribed to Garrett: Hume is here denying that such inferences can be explained as an activity of the faculty of reason conceived as functioning by the discovery of intermediate ideas . . . Garrett says that Hume ‘is denying only that we come to engage in this species of reasoning as a result of any piece of reasoning about it’.72. . . My main objection to Garrett’s interpretation is that he treats Hume as asking about the cause of our engaging in probable reasoning . . . Hume’s question is not what Garrett takes it to be. Hume’s question is: how is it that we manage to make these inferences?73

Hume produced no fewer than seven editions of his Enquiry after Price published this, and it would be astonishing if he did not know of it, given that the Enquiry itself – under its original title Philosophical Essays – is mentioned twice within the vicinity of these quotations (in footnotes on pages 12 and 41). As interpreted by Owen and Garrett, however, Hume’s famous argument completely fails to engage with Price’s justification of induction. That justification starts from what Hume would no doubt claim to be a fallacious ‘intuition’, but again there is no obvious psychological obstacle to erecting an argument on a fallacy: humans do it all the time! Another related issue concerns the intended scope of the famous argument: is it supposed to be proving something about every individual factual inference, or about the genesis of our general practice of factual inference? When developing his interpretation, Owen writes repeatedly as though it were the latter, for example:

As we have seen already, however, Garrett’s interpretation in his 1997 book was quickly modified (in his 1998 debate with me), at which point he clarified that, like Owen, he took Hume’s conclusion to be one that applies to all individual inductive inferences: [From] Cognition and Commitment . . . , Millican understandably infers that on my interpretation ‘it is only the general practice of induction that fails to be determined by reason, and each of our particular inductive inferences is itself an instance of the operation of our reason.’ . . . The crucial distinction for Hume, however, is . . . between an inference being an instance of reasoning and the same inference being caused by (another instance of) reasoning.74

If the uniformity principle were something we knew or believed, prior to our engaging in probable reasoning,70 we could explain probable reasoning as being based on reason. . . . [P]rior to our engaging in probable reasoning, we . . . neither know nor believe the uniformity principle. . . . Hume’s argument . . . has more to do with the failure of reason to account for why we have the beliefs 77

9780826443595_Ch03_Final_txt_print.indd 77

2/9/2012 9:38:52 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION of probable reasoning, it will involve a crucial step that is not supported by any argument . . . ; that is the point of Hume’s discovery. Reasoning cannot cause the crossing of an inductive gap.77

The significance of this point is that the bulk of Hume’s argument, and especially his conclusion – in both the Treatise and the Enquiry – seems to intend a result about every particular factual inference (cf. section 2.4 [H] above). But on the interpretations of Owen and Garrett, the argument seems to lead far more plausibly to a result about factual inference in general. Note, for example, that Hume takes many inductive inferences to be reflective and mediated, especially those that involve ‘inferences from contrary phænomena’ (THN 1.3.12.7 / 133) or the application and balancing of ‘general rules’ (THN 1.3.13.7–12 / 146–50, 1.3.15 / 173–5). Moreover, inductive inferences from ‘only one experiment of a particular effect’ can – Hume says – be mediated by explicit reflection on a principle which looks very similar to UP (THN 1.3.8.14 / 104–5).75 Owen is well aware of this,76 but he does not apparently recognize the threat to his own interpretation, under which such mediated inferences become counterexamples to Hume’s conclusion, at least if that is read (as Hume’s own words seem to require) as a claim about each and every factual inference. Garrett is also aware that ‘not all probable inferences are immediate’, but he endeavours to explain how nevertheless Hume’s conclusion can be seen to apply even to those that are mediated:

However, my objection to which this was a response referred not to the situation where ‘one piece of probable reasoning is part of another piece of probable reasoning’, but rather, where one inductive inference’s conclusion (e.g. ‘a general principle of uniformity’ as at THN 1.3.8.14 / 104–5) is then given ‘the role of a premise in further inductive inference’.78 Garrett’s response thus preserves what he takes to be Hume’s conclusion, as universally applicable to factual inferences, only by stipulatively treating the inference which was used to establish a proposition as itself a part of the further inference which then takes that proposition as a premise. In the context of discussing the epistemology of induction, this might seem reasonable enough: if the proposition in question has a problematic foundation, then those problems will be inherited by any further inference built on it. But if Hume’s famous argument is to be interpreted as involving the psychological mechanism of individual inductive inferences – as Garrett intends – then the move looks artificial and ad hoc, smudging over the manifest difference between arguing for some proposition and taking it as assumed or established (on the basis of previous argument). To sum up so far, we have yet to find an interpretation that is genuinely satisfying. Any extreme sceptical reading leaves Hume’s philosophy hopelessly inconsistent. The anti-deductivist reading and that of Owen both have serious difficulty making sense of the logic of his argument, and also have to rely on the textually questionable claim that Hume’s notion of

it may well happen, as Millican notes, that one piece of probable reasoning is part of another piece of probable reasoning . . . But as Hume states his conclusion . . . , it is that ‘in all reasonings from experience, there is a step taken by the mind which is not supported by any argument or process of the understanding’ [EHU 5.2 / 41; emphasis added]. Where a piece of probable reasoning does occur as part of a second piece 78

9780826443595_Ch03_Final_txt_print.indd 78

2/9/2012 9:38:53 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION is looking far more like a discussion of the epistemology of our general presumption of uniformity.

‘reason’ within the argument is a target rather than sincere (a problem which also beset my own previous interpretation).79 Garrett’s reading has the significant merit of avoiding this last pitfall, but again has difficulty squaring with the argument’s text and logic, and has been forced to adapt accordingly over time. Initially, Garrett understood Hume’s conclusion as the straightforward claim that ‘we do not adopt induction on the basis of recognizing an argument for its reliability’.80 This soon changed into the more complex claim that ‘the underlying causal mechanism that gives rise to inductive inferences . . . is not itself dependent on any reasoning or inference’.81 Meanwhile, since the Enquiry argument clearly ranges beyond narrow ‘reasoning or inference’, Garrett suggested that here Hume ‘expands the famous conclusion to rule out any “reasoning or process of the understanding,” thereby eliminating such non-inferential processes of the understanding as intuition or the perception of a probable connection between even a single “proof” and a conclusion’.82 But pushing in the opposite direction, his recognition that Hume acknowledges the role of explicit (and sometimes complex) ratiocination within some inductive inferences has led to a narrowing of the supposed conclusion, to focus on the very specific logical step which is ‘the crossing of an inductive gap’.83 Even after all this, as we have seen, Garrett’s defence of the interpretation looks suspiciously ad hoc, holding that conclusion to be true even of an inductive inference which explicitly argues across the inductive gap using an antecedently established Uniformity Principle, simply on the basis that some previous inference was required to establish that principle. But by now the interpretation has been diluted beyond recognition, and we seem to have lost any focus on the actual psychological mechanism of individual inductive inferences – this

3.2 ‘REASON’ AS THE COGNITIVE FACULTY One promising route towards a better understanding of Humean ‘reason’ is to look at the usage of Hume’s contemporaries in Scotland and England, and especially those who – unlike Locke – were enthusiastic about the language of ‘faculties’ and relatively consistent in their usage.84 Francis Hutcheson, Professor of Moral Philosophy at Glasgow and correspondent with Hume from 1739, provides the closest spatio-temporal match to the Treatise, having published in 1742 no fewer than four works containing an outline of the faculties, at least one of which – Philosophiae Moralis Institutio Compendiaria (later translated as A Short Introduction to Moral Philosophy) – he sent to Hume.85 This, like the Synopsis Metaphysicae (Synopsis of Metaphysics)86 was a Latin teaching text, and contains a discussion of ‘The parts or powers of the soul’.87 The other two works were An Essay on the Nature and Conduct of the Affections and Illustrations on the Moral Sense, published together as a third edition of both. To the former, Hutheson added a footnote on the faculties,88 and to the latter, a new paragraph: Writers on these Subjects should remember the common Divisions of the Faculties of the Soul. That there is 1. Reason presenting the natures and relations of things, antecedently to any Act of Will or Desire: 2. The Will, . . . or the disposition of Soul to pursue what is presented as good, and to shun Evil. . . . Both these Powers are by the Antients included under the Λόγος or λογικòν μέρος. Below these they place 79

9780826443595_Ch03_Final_txt_print.indd 79

2/9/2012 9:38:53 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION This sceptical doubt, both with respect to reason and the senses . . . ’Tis impossible upon any system to defend either our understanding or senses . . . ’ (THN 1.4.2.57 / 218; emphasis added)

two other powers dependent on the Body, the Sensus, and the Appetitus Sensitivus, in which they place the particular Passions: the former answers to the Understanding, and the latter to the Will. But the Will is forgot of late, and some ascribe to the Intellect, not only Contemplation or Knowledge, but Choice, Desire, Prosecuting, Loving.89

Consider also the following two footnotes, the first of which was expanded and moved to create the second: when it [the imagination] is oppos’d to the understanding, I understand the same faculty, excluding only our demonstrative and probable reasonings. (THN 2.2.7.6n / 371n; emphasis added)

It is clear from his alternation between ‘Reason’ and ‘the Understanding’ that Hutcheson takes these to be one and the same; indeed his Essay’s new footnote says as much.90 This equivalence is also asserted (or manifested through the same sort of elegant variation of terminology that we see above) by various other writers known to Hume, for example Shaftesbury, Butler and Price,91 so it was evidently commonplace, though writers in the Scottish common-sense school later preferred to use ‘reason’ more narrowly, in much the way that Garrett favours.92 Hume himself, however, is clearly aligned with the former group, interchanging between ‘reason’ and ‘the understanding’ dozens of times – purely for the sake of stylistic variation – just as he does between ‘the fancy’ and ‘the imagination’:93

when I oppose it [the imagination] to reason, I mean the same faculty, excluding only our demonstrative and probable reasonings. (THN 1.3.9.19n22 / 117–18n; emphasis added)

Again, the switch from ‘the understanding’ to ‘reason’ looks purely stylistic, perhaps prompted by the clumsiness of ‘. . . the understanding, I understand . . . ’. There is thus overwhelming textual evidence that Hume generally treats ‘reason’ and ‘the understanding’ as one and the same. And virtually all major writers of the period take ‘the understanding’ to refer to our principal cognitive faculty, usually drawing a general division between it and ‘the will’.94 This division between the domains of the understanding and the will is indeed essentially the same as the modern distinction between cognitive and conative mental functions, a dichotomy whose fundamental nature is often now expressed in terms of a ‘direction of fit’ between world and mind: the understanding aims to conform our beliefs to the way the world is, while the will aims to change the world to conform to our desires. Reid characterizes this in terms of a

the mind . . . is not determin’d by reason, but by certain principles, which associate together the ideas of these objects, and unite them in the imagination. Had ideas no more union in the fancy than objects seem to have to the understanding, . . . (THN 1.3.6.12 / 92; emphasis added) There are no principles either of the understanding or fancy, which lead us directly to embrace this opinion . . . The . . . hypothesis has no primary recommendation either to reason or the imagination . . . (THN 1.4.2.46 / 211; my emphasis) 80

9780826443595_Ch03_Final_txt_print.indd 80

2/9/2012 9:38:53 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION distinction between our intellectual (or contemplative) and active powers:

This treats the senses themselves as ‘operations of the understanding’, a tendency common enough for Price to make a point of criticizing it.98 But Hutcheson’s Synopsis of Metaphysics, within two sentences, first implicitly places the senses within the understanding and then gives them a subordinate reporting role, which suggests that the former placement is just a shorthand way of indicating that the senses fall within the understanding’s sphere of influence:

We shall . . . take that general division which is the most common, into the powers of understanding and those of will. Under the will we comprehend our active powers, and all that lead to action, or influence the mind to act; such as, appetites, passions, affections. The understanding comprehends our contemplative powers; by which we perceive objects; by which we conceive or remember them; by which we analyse or compound them; and by which we judge and reason concerning them. . . . The intellectual powers are commonly divided into simple apprehension, judgment, and reasoning.95

we might reasonably reduce [the powers of the mind] to two, namely, the faculty of understanding and the faculty of willing, which are concerned respectively with knowing things and with rendering life happy. The senses report to the understanding, . . .99

Although Reid is somewhat critical of this framework, which he takes to be ‘of a very general reception’, his clear account of it is helpful in setting the scene for Hume, whose understanding of it – though again critical – seems to be very similar.96 The earlier quotation from Hutcheson’s Illustrations on the Moral Sense likewise recognizes this ‘common Division of the Faculties of the Soul’ between ‘Reason’ (or ‘the Understanding’) and ‘The Will’, while suggesting a hierarchical structure in which the senses ‘answer to’ the understanding, and the passions to the will. His Short Introduction to Moral Philosophy, however, paints a cruder picture which is closer to that outlined by Reid above:

The Synopsis goes on to give Hutcheson’s most detailed account of his faculty framework, with Chapter 1 of Part II devoted to a categorization of the powers associated with the understanding, including external sensation (sect. 3), internal senses or consciousness (sect. 4), reflexive or subsequent sensations (sect. 5), memory, the power of reasoning and imagination (sect. 6).100 This again suggests a hierarchical structure, with these various powers ‘reporting to’ an overseer faculty – reason or the understanding proper – which perceives and judges the deliverances of the subordinate faculties in order to establish truth. Thus ‘Reason is understood to denote our Power of finding out true Propositions’.101 Price talks in a similar spirit of ‘the power within us that understands; . . . the faculty . . . that discerns truth, that views, compares, and judges of all ideas and things’.102 And a similar conceptual linkage between reason or the understanding and the search for truth is common to many other writers.103

The parts or powers of the soul . . . are all reducible to two classes, the Understanding and the Will. The former contains all the powers which aim at knowledge; the other all our desires. . . . [Of] the several operations of the understanding . . . The first in order are the senses . . . Senses are either external, or internal.97 81

9780826443595_Ch03_Final_txt_print.indd 81

2/9/2012 9:38:53 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

That faculty, by which we discern Truth and Falshood, . . . (EHU 1.4n – 1748/1750 editions);105

is entirely unreasonable, must proceed from some other faculty than the understanding. . . . Even after we distinguish our perceptions from our objects, ’twill appear presently, that we are still incapable of reasoning from the existence of one to that of the other: So that upon the whole our reason neither does, nor is it possible it ever shou’d, . . . give us an assurance of the continu’d and distinct existence of body. (THN 1.4.2.14 / 193)

Thus the distinct boundaries and offices of reason and of taste are easily ascertained. The former conveys the knowledge of truth and falsehood: . . . (EPM App. 1.21 / 294);

’tis a false opinion that any of our . . . perceptions, are identically the same after an interruption; and consequently . . . can never arise from reason, . . . (THN 1.4.2.43 / 209)

We are now in a position to appreciate the significance of Hume’s repeated statements that align him strongly with this general conception of reason as cognition:104 Reason is the discovery of truth or falshood. (THN 3.1.1.9 / 458);

we may observe a conjunction or a relation of cause and effect betwixt different perceptions, but can never observe it betwixt perceptions and objects. ’Tis impossible, therefore, that from the existence or any of the qualities of the former, we can ever form any conclusion concerning the existence of the latter,106 or ever satisfy our reason in this particular. (THN 1.4.2.47 / 212)

reason, in a strict sense, as meaning the judgment of truth and falsehood, . . . (DIS 5.1 / 24; cf. THN 2.3.3.8 / 417). In all these contexts, Hume is stressing that reason, since it is purely cognitive, cannot also be conative: that is, it cannot be contrary to any passion or – by itself – provide any motive to action or the will. This is the crux of one of Hume’s three most famous arguments concerning the incapacity of reason, which concludes that ‘Since morals . . . have an influence on the actions and affections, it follows, that they cannot be deriv’d from reason’ (THN 3.1.1.6 / 457). For the purpose of this argument, it is enough that reason is confined to the domain of truth and falsehood, though Hume’s talk of discovery, discernment and knowledge suggests a normative bias towards truth rather than falsehood. This normative flavour is far more explicit in another of the three famous arguments, this time concerning the external world (the third, of course, concerns induction):

It might appear strange that in one of these arguments, reason seems to embrace both truth and falsehood, whereas in the other, it is normatively connected with truth. But this sort of linguistic variation is commonplace, and it is worth noting that an unambiguous identification of reason with the cognitive faculty is consistent with a fairly wide range of nuances of meaning. Given such an identification, ‘reason’ might most naturally be used to refer to the human (or animal) faculty of truth-apprehension, however well and by whatever processes it operates (as, for example, when Hume compares the ‘reason’ of people and animals at THN 3.3.4.5 / 610). But sometimes there might be debate over these processes, in which case we could

The vulgar confound perceptions and objects . . . This sentiment, then, as it 82

9780826443595_Ch03_Final_txt_print.indd 82

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION our natural faculties, then one would expect that his arguments about reason’s capabilities would start from a relatively straightforward and conventional understanding of our cognitive functions. Coming from a Lockean background, it is no surprise to find Hume recognizing the cognitive faculties of the senses, memory, intuition, and demonstrative and probable reasoning.109 The senses can be either external (i.e. sight, touch, hearing, smell, gustatory taste) or internal (i.e. reflection) – these provide the impressions from which our ideas are copied, and those ideas are represented to us either through the memory or the imagination. It follows that all of our thinking, except in so far as it confines itself to memory, must involve representation of ideas in the imagination, which is apparently to be thought of as something like a multi-layered or multi-dimensional canvas on which sense-copied ideas appear, with different degrees of ‘force and vivacity’ and ‘in a perpetual flux and movement’ (THN 1.4.6.4 / 252).110 Thus our faculties of intuition, demonstration, and probable reasoning must inevitably act on our imagination, through such processes as bringing ideas into mind, dismissing others, or – most importantly given Hume’s analysis of belief (summarized at the end of section 2.1 above) – changing their degrees of force and vivacity. Even when we make judgements about the deliverances of our senses and memory, it is their force and vivacity in the imagination which apparently constitutes our assent to them.111 Hence Hume’s comment, leading into the sceptical anxieties of the Conclusion of Treatise Book 1, that: ‘The memory, senses, and understanding are, therefore, all of them founded on the imagination, or the vivacity of ideas.’ (THN 1.4.7.3 / 265). This comment might be read as suggesting that the imagination is itself active, but earlier in the same paragraph, Hume

find ourselves referring to processes that are commonly taken to be involved in truthapprehension, even if they turn out not to be truth-conducive (as suggested by Hume’s ‘scepticism with regard to reason’ of Treatise 1.4.1). Alternatively, we might wish to apply a stricter criterion under which ‘reason’ would be confined to processes that operate successfully to apprehend truth (thus giving the normative flavour of the passages from Treatise 1.4.2 quoted above).107 A different strict usage is to refer to the faculty of truth-apprehension acting entirely alone, independently of other faculties such as the senses or memory (this seems to be Hume’s intention at THN 3.1.2.1 / 470).108 Finally, there is in the early modern period a common metonymy, under which ‘reason’ is used to refer to its product, namely true belief as successfully achieved using our rational faculty (hence the pairing of ‘truth and reason’ at THN 2.3.3.5 / 415, 3.1.1.15 / 461, and THN App. 1 / 623; cf. also note 84 above). Notice that acknowledgement of all these nuances is quite different from supposing an ambiguity in ‘reason’, because they all arise naturally from the core meaning, and there need be no suggestion that the word has been coincidentally assigned two or more distinct meanings. With this understood, much of the evidence that has previously been adduced for the ambiguity of ‘reason’ is significantly undermined, and it becomes more plausible to suggest that the term has, for Hume, a single core meaning, namely what we now call cognition. 3.3 REASON AND THE IMAGINATION If reason, for Hume, is just our overall cognitive faculty, and if his general epistemological approach is – as set out in section 1 above – to begin by ascribing default authority to 83

9780826443595_Ch03_Final_txt_print.indd 83

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION the opprobrious character of being the offspring of the imagination. By this expression it appears that the word, imagination, is commonly us’d in two different senses; and tho’ nothing be more contrary to true philosophy, than this inaccuracy, yet in the following reasonings I have often been oblig’d to fall into it. When I oppose the imagination to the memory, I mean the faculty, by which we form our fainter ideas. When I oppose it to reason,115 I mean the same faculty, excluding only our demonstrative and probable reasonings. (THN 1.3.9.19n22 / 117–18n)

makes clear that he is talking of principles (namely experience and habit) that ‘operate upon the imagination’. The initial framing of his discussion of induction in the Treatise (as quoted earlier in section 2.1) may give a different impression: ‘the next question is, Whether experience produces the idea by means of the understanding or imagination; whether we are determin’d by reason to make the transition, or by a certain association and relation of perceptions?’ (THN 1.3.6.4 / 88–9). But Hume’s answer to his own question – repeated numerous times – will be that our causal reasoning is determined by custom,112 and he never says that it is determined by the imagination itself. So at least in this context, the imagination is apparently only the virtual stage on which the mind’s various principles – either of reason or custom – orchestrate their dance of perceptions.113 In other contexts, however, the imagination does appear as an active agent, having the liberty to transpose and change its ideas (THN 1.1.3.4 / 10, 1.1.4.1 / 10–11, 1.3.5.3 / 85, 1.3.7.7 / 97; EHU 2.4 / 18, 5.10 / 47–8, 5.12 / 49), to distinguish and separate them (THN 1.2.4.3 / 40, 1.2.5.3 / 54–5, 1.3.3.3 / 79–80, 1.4.5.5 / 233), to suggest (THN 1.1.7.15 / 23–4) or raise them up (THN 1.2.1.3 / 27), and to generate fictions (THN 1.1.6.2 / 16, 1.4.2.29 / 200–1, 1.4.2.36 / 205, 1.4.2.43 / 209, 1.4.2.52 / 215, 1.4.3.3–5 / 220–1, 1.4.3.11 / 224–5, 1.4.6.6–7 / 253–5).114 The distinction between the two classes of operation seems to be explained by the footnote at THN 1.3.9.19 / 117 which we encountered in section 3.2 above:

This note was inserted by means of a ‘cancel’ leaf, prepared by Hume while the Treatise was going through the press, and I believe he saw the need for this on rereading the end of THN 1.3.9.4 / 108:116 ‘All this, and every thing else, which I believe, are nothing but ideas; tho’ by their force and settled order, arising from custom and the relation of cause and effect, they distinguish themselves from the other ideas, which are merely the offspring of the imagination.’ A related passage is at THN 1.4.4.1 / 225, where Hume addresses the complaint that he has criticized ‘the antient philosophers’ for being guided by imaginative fancies, whilst building his own philosophy on principles of the imagination: In order to justify myself, I must distinguish in the imagination betwixt the principles which are permanent, irresistible, and universal; such as the customary transition from causes to effects, and from effects to causes: And the principles, which are changeable, weak, and irregular; such as those I have just now taken notice of [e.g. the ‘inclination in human nature to bestow on external objects the same emotions, which it observes in itself’, as attributed to the ‘antient

In general we may observe, that as our assent to all probable reasonings is founded on the vivacity of ideas, it resembles many of those whimsies and prejudices, which are rejected under 84

9780826443595_Ch03_Final_txt_print.indd 84

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION philosophers’]. The former are the foundation of all our thoughts and actions, so that upon their removal human nature must immediately perish and go to ruin. The latter are neither unavoidable to mankind, nor necessary, or so much as useful in the conduct of life . . .

our memory’ (EHU 4.3 / 26). He then identifies the crucial step of such inference: the extrapolation from observed to unobserved which is encapsulated in his Uniformity Principle. If this is to qualify as founded on reason, then there must be some cognitive operation that grounds it, and which does so through genuine cognition (rather than some fallacy or confusion). In the Treatise and Abstract, Hume apparently takes it to be obvious that the only plausible contenders here are demonstrative reasoning and probable (i.e. moral or factual) reasoning. In the Enquiry he is more thorough, and rules out also both intuition and sensory knowledge as sources of foundation for the Uniformity Principle. Since memory is taken for granted in the experiential observations from which the inference starts, this exhausts all the standardly recognized sources of evidence with which reason might operate. It is therefore no coincidence that the four sources considered – and rejected – in Hume’s argument in the 1748 Enquiry match up exactly with those itemized in his 1745 Letter from a Gentleman: ‘It is common for Philosophers to distinguish the Kinds of Evidence into intuitive, demonstrative, sensible, and moral; . . .’ (LFG 22). If reason is understood by Hume in the standard contemporary way – as the overall cognitive faculty – then we should indeed expect it to embrace all four ‘Kinds of Evidence’. Notice that this way of reading Hume’s argument has the clear implication that inductive (i.e. probable, moral, factual) inference is being treated as an operation of reason throughout, which at least strongly suggests that it would be a mistake to interpret Hume’s conclusion – that such inference is ‘not determin’d by reason’ – as deposing it from that status. (For if that were indeed his intention, one might reasonably expect

All three passages point to a distinction drawn within the class of principles that ‘operate on the imagination’ – that is, which affect our thinking. Some of these are ‘the foundation of all our thoughts and actions’, the ‘permanent, irresistible, and universal’ principles that ground ‘our demonstrative and probable reasonings’, and are therefore appropriately dignified with the name of reason or the understanding. The other principles are those that we more naturally think of as belonging to the imagination itself: those that ground our free play of ideas, fictions, whimsies and prejudices. Hence in this narrower sense the imagination is in opposition to reason, though both sets of principles perform on the same stage – the imagination in the broader sense – where all our non-memory ideas are represented. It is this broader sense which enables Hume to refer, without paradox, to ‘the understanding, that is, . . . the general and more establish’d properties of the imagination’ (THN 1.4.7.7 / 267).117 3.4 AN OPERATION OF REASON WHICH IS ‘NOT DETERMIN’D BY REASON’ Equipped with this understanding of the Humean faculties, let us now try to clarify the significance of his famous argument. In the Treatise, he considers a paradigm causal inference ‘from the impression to the idea’; in the Abstract and Enquiry, he widens this to any factual inference ‘beyond the present testimony of our senses, or the records of 85

9780826443595_Ch03_Final_txt_print.indd 85

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION A similar theme can be seen in the other of Hume’s most famous arguments that assigns a vital role to the imagination, on ‘Scepticism with Regard to the Senses’ (THN 1.4.2 / 187–218). Here he takes on the natural and naive assumption that external objects – distinct from us and continuous over time – are directly and straightforwardly perceived through the senses. To refute this, he shows that identification of objects over time requires a process that goes beyond anything we perceive, latching onto patterns of ‘constancy’ and ‘coherence’ in our distinct impressions, and smoothing over gaps and changes to generate an illusion of continuity. Again the process involved is naturally categorized as ‘imaginative’, and so Hume describes his argument as showing that our ‘assurance of the continued and distinct existence of body . . . must be entirely owing to the imagination’ (THN 1.4.2.14 / 193). Like his argument concerning induction, therefore, this can be seen as making a significant contribution to both cognitive science and epistemology, by highlighting how the informational processes that are implicit in the temporal identification of physical objects go well beyond anything that is directly perceived. Indeed modern-day cognitive science, through the development of ‘artificial intelligence’ visual systems, has provided striking vindication of Hume, by showing how even the identification of physical objects at a time requires ‘imagination-like’ processes of edgedetection, region identification, shadow and texture interpretation, and so forth. So far from being merely passive, visual perception involves many active – albeit unconscious – processes, without which the manifold of our sensory impressions would be completely incomprehensible. This interpretation of Hume involves understanding his talk of faculties as

such an apparently paradoxical move to be far more clearly signalled.) So we need to understand how Hume, as a result of his famous argument, can coherently view induction as an operation of reason which is not ‘determin’d by’ reason.118 The obvious answer, given both our interpretation of reason and the structure of Hume’s argument, is that he views induction as a cognitive process which depends on a non-cognitive sub-process. So he is thinking at two levels, with inductive inference being a manifest operation of our conscious reason, causally driven by a subconscious process that involves the customary enlivenment of our ideas. This underlying process is of a type which is naturally categorized as ‘imaginative’ rather than ‘rational’, because it works through an associative mechanism which automatically and mindlessly extrapolates beyond anything that we have perceived or otherwise detected in the world (whether objective events, or evidence). It is therefore in sharp contrast with the underlying process hypothesized by Locke, who supposed inductive (i.e. probable) inference to be driven by a perceptual process, namely the rational apprehension of objective probable connexions. Locke therefore saw induction as a cognitive process which depends on a cognitive sub-process – apparently ‘cognitive all the way down’ because it is ultimately founded on direct perception of evidential connexions. Hume’s argument destroys this illusion by showing that there is no plausible source for such perception: it cannot derive from examining relations of ideas (because it depends on the experienced world), but nor can it derive from experience, either current (because the senses detect no such evidential connexions) or remembered (because inductive extrapolation is the very process whose perceptual basis we are seeking). 86

9780826443595_Ch03_Final_txt_print.indd 86

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION stress the liberty of the imagination (EHU 2.4 / 18, 5.10 / 47–8, and 5.12 / 48–50). The contrast is especially striking in the case of reason, because whereas the Treatise speaks of reason itself as a determining cause (e.g. THN 1.3.6.4 / 88–9, 1.3.6.12 / 92, 1.3.7.6 / 97, 1.4.1.1 / 180), the Enquiry never does so. In the later work, Hume’s preference is to talk instead of reasoning processes (e.g. EHU 4.23 / 39, 5.4 / 42, 9.6 / 108), which were never mentioned as such in the Treatise. Meanwhile, custom in the Enquiry is said to act on the imagination (EHU 5.11 / 48, 9.5 / 106–7) and is never said (or implied) to be itself an operation of the imagination, thus avoiding the complications that arise from trying to place it consistently within the conventional faculty structure.122 Moreover, Hume no longer refers to custom as a principle of association of ideas (cf. THN 1.3.7.6 / 97),123 but says instead that it is a process analogous to the association of ideas, which ‘is of a similar nature, and arises from similar causes’ (EHU 5.20 / 53–4). He continues, however, to draw a contrast between custom and reason (EHU 5.5 / 43, 5.20 / 53–4), thereby retaining the core of his theory that inductive inference is determined by a subprocess which is not itself cognitive.

descriptive of types of process rather than as references to parts of the mind, and indeed this seems anyway to be required in the light of his general scepticism about any faculty language that pretends to be more than a functional description (THN 1.4.3.10 / 224; DNR 4.162–3). For Hume, as for Locke, a faculty just names a power.119 Nevertheless, at least in the Treatise, he has an unfortunate tendency to talk of faculties in the way that Locke rightly deplored, as ‘so many distinct Agents’.120 Taking such language literally, his famous argument paints the absurd picture of reason attempting in vain to make an inductive inference, and needing the imagination to step in to lend a hand. But induction is such a central cognitive process that it ought by definition to be an operation of reason (just as remembering is by definition an operation of memory); hence if we think of faculties as distinct agents or areas of the mind, custom – as the underlying process that drives induction – should itself be part of reason. Presumably it is this sort of consideration that led Hume, in the wake of his famous argument, to reassign ‘the general and more establish’d properties of the imagination’ – which must surely include custom – to reason or the understanding (THN 1.4.7.7 / 267, cf. section 3.3 above).121 But then we get into a muddle if we want to hold on to his conclusion that inductive inference is ‘not determin’d by reason’, given the frequency with which he says that inductive inference is indeed determined by custom. Little wonder, perhaps, that both Hume and his interpreters sometimes seem to exhibit confusion of the faculties! As so often, the Enquiry brings considerable improvement, and in a number of respects. Now the faculties are rarely spoken of as agents in their own right, with the harmless exception of those passages that

4. CONCLUSION: SCEPTICISM AND RATIONAL FOUNDATIONS After all this, how sceptical is Hume’s position? His famous argument has shown that inference from past to future crucially involves a process of extrapolation that cannot be independently justified by anything within our cognitive grasp. This crucial step is instead due to a mechanical associative process in the mind, whereby past experience 87

9780826443595_Ch03_Final_txt_print.indd 87

2/9/2012 9:38:54 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION of impressions), he expresses his conclusion in a way that ignores the obvious and essential role of the senses:

raises certain ideas about the future and enlivens them into beliefs. Such a process – given its automatic, non-reflective nature, and its lack of any rational insight or apprehension of reality – is naturally classified as ‘imaginative’ rather than ‘cognitive’, and Hume’s faculty language is best interpreted accordingly, as a way of categorizing types of process, rather than as a theory of distinct agents within our minds. So when he claims that the imagination plays a crucial role in inductive inference, he should be understood as saying simply that our process of making inductive inferences itself crucially involves an imagination-like sub-process. As we have seen (in section 3.1 above), Hume is well aware that many inductive inferences also involve reason-like sub-processes, as for example when we consciously take into account the ‘rules by which to judge of causes and effects’ of Treatise 1.3.15 / 173–6, or attempt to identify underlying mathematical patterns (e.g. EHU 4.13 / 31, 7.29n17 / 77n). But he is clearly far more interested in the crucial imaginative step, even to the extent of describing it as solely responsible for the inference:

That opinion must be entirely owing to the imagination (THN 1.4.2.14 / 193; emphasis added). Hume seems to be assuming here that even one imaginative step is sufficient to characterize the entire process of which it is a part as one that is determined by the imagination (rather than by reason or the senses), just as one invalid step within a sequential inference typically renders the entire inference invalid. Such a focus when speaking of inferential processes is indeed quite natural, since we are typically interested in the weakest link in any chain of support rather than the strongest. The same applies to other supportive or foundational relationships: thus a climber can properly be described as ‘supported only by a rope’, whether that rope itself is secured to a mountain, a building, a heavy vehicle or any other relatively reliable anchor.124 Likewise, an argument or legal case which crucially depends on some imaginative fabrication, even if it also depends on numerous points that are logically unassailable, can appropriately be said to be ‘founded on fantasy’. But if we follow through this line of thought, then since inductive inference depends on a sub-process of ‘imaginative’ extrapolation which itself has no rational grounding, we seem forced to conclude that any proposition that can be established only by such inference must apparently in turn be disqualified from counting as founded on reason. Yet as we have seen in sections 3.1 and 3.4 above, Hume continues to treat induction as a legitimate operation of reason.125 There is, at the least, a sceptical tension here: can we really suppose that he would consider a process

When the mind . . . passes from the idea or impression of one object to the idea or belief of another, it is not determin’d by reason, . . . The inference . . . depends solely on the union of ideas. (THN 1.3.6.12 / 92; emphasis added) all reasonings are nothing but the effects of custom; . . . (THN 1.3.13.11 / 149; emphasis added; cf. EHU 5.5–6 / 43–4) Similarly with his argument concerning our belief in the continued and distinct existence of body (which aims to show that it depends crucially on various associative processes, constructing ‘fictions’ from the passing show 88

9780826443595_Ch03_Final_txt_print.indd 88

2/9/2012 9:38:55 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION foundation for belief and also a cause. By contrast, ‘imagination-like’ processes such as custom may cause belief, but they cannot at the same time provide a cognitive foundation: that is indeed precisely why they do not qualify as processes of reason.127 Hume seems to have embraced this distinction, if not perhaps immediately, for his language in the Abstract and Enquiry (though not in the Treatise) precisely fits it. In the Treatise, he repeatedly talks about custom (or the principles of the imagination) as providing a foundation for inductive inference.128 In the Abstract and Enquiry, by contrast, he never does, but there are no fewer than 19 passages that describe the influence of custom in terms that are either explicitly causal, or naturally interpretable as such.129 This strongly suggests that Hume himself came to recognize a firm distinction between what in the Enquiry he calls a foundation, and what he there calls a determining cause. Thus understood, it is clear that custom qualifies only as a cause, whereas reasoning processes or sources of information can potentially provide a cognitive foundation.130 All this brings the possibility of posing coherent but unanswerable questions, such as that which introduces Hume’s discussion in Part 2 of Enquiry Section 4: ‘if we still carry on our sifting humour, and ask, What is the foundation of all conclusions from experience? this implies a new question . . .’ (EHU 4.14 / 32). If custom cannot qualify as a foundation, then Hume’s ultimate conclusion that ‘All inferences from experience . . . are effects of custom, not of reasoning’ (EHU 5.5 / 43) excludes any foundation at all. For in competing successfully for the causal explanatory role, custom effectively excludes anything else from the foundational role (which it is nevertheless unable to fulfil itself). Perhaps, then, there is hidden depth

genuinely rational which rests on a purely mechanical, non-reflective sub-process? To address this worry, suppose that Hume were to take the alternative view, that any rational process must have a rational foundation. It would then immediately follow that for anything to be founded on reason at all, it must be founded on reason ‘all the way down’ (i.e. it would have to be solidly founded on evidence or principles, which are either immediately apprehended by reason, or else themselves solidly founded on evidence or principles which are either immediately apprehended by reason or . . . , etc.). Hume would thus be committed to a strongly rationalistic notion of reason, the demands of which would be impossible to fulfil without abandoning the heart of his philosophy. At no point would he be able to halt the foundational regress by acknowledging that ultimately the principles of our reason can (legitimately) be grounded on basic psychological mechanisms. So the only possible outcomes would be either rationalism or incurable scepticism. Some interpreters have indeed seen Hume as impelled towards radical scepticism by precisely this kind of regressive train of thought.126 But it would be completely at odds with his efforts to ground a conception of reason on the contingent operations of the human mind, and flatly incompatible – in the light of his own investigations – with treating induction as a genuine operation of reason. As we saw in section 3.4 above, Locke implicitly follows the path that Hume rejects, by attributing probable reasoning to the perception of probable connexions. And indeed direct perception – conceived of as a process of transparent apprehension – seems to be a paradigm of what reason requires if it is to be ‘cognitive all the way down’. Such perception could at once provide both a rational 89

9780826443595_Ch03_Final_txt_print.indd 89

2/9/2012 9:38:55 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION inference possible for human beings, despite the sceptical impact of his famous argument which shows that it cannot be founded in any of the ways that previous tradition would countenance.131 He reveals how we actually reason inductively, rather than falling back on the aprioristic supposition that this can only be through the rational perception of evidential connexions. That traditional notion is decisively refuted by his sceptical argument, but his own position is very far from sceptical. On the contrary, as we saw in section 1 above, Hume sees very good reason to accept our faculty of inductive inference as it is (at least when suitably disciplined by general rules etc.), and no good reason to reject it. We have, indeed, no alternative, nor any compelling reason for desiring one.132

in Hume’s declaration of intent: ‘I shall content myself, in this section, with an easy task, and shall pretend only to give a negative answer to the question here proposed.’ (EHU 4.15 / 32). The upshot is that ‘if we still carry on our sifting humour’ in the search for ultimate foundations, we hit rock bottom with something that has a cause but no foundation. And that is the tendency, rooted in our animal nature, to infer from past to future, from experienced to not-yet-experienced. This is radically different from the kind of perceptual foundation presupposed by traditional conceptions of reason, different enough to make Hume’s position seem outrageously sceptical by comparison. But in reality it is quite the reverse: he is providing an account of reason which makes inductive

APPENDIX: HUME’S ARGUMENT CONCERNING INDUCTION (FROM SECTION 4 OF THE ENQUIRY CONCERNING HUMAN UNDERSTANDING) It is allowed on all hands, that there is no known connexion between the sensible qualities and the secret powers . . . (EHU 4.16 / 33) 4. . . . every effect is a distinct event from its cause. It could not, therefore, be discovered in the cause, and . . . the conjunction of it with the cause must appear . . . arbitrary; since there are always many other effects, which, to reason, must seem fully as consistent and natural. (EHU 4.11 / 30) 5. . . . the knowledge of [cause and effect] is not, in any instance, attained by reasonings a priori; but arises entirely from experience . . . (EHU 4.6 / 27) . . . causes and effects are discoverable, not by reason, but by experience . . . (EHU 4.7 / 28) In vain, therefore, should we pretend to . . . infer any cause or effect, without

Hume’s Own Statement of the Propositions Identified in the Structure Diagram 1. By means of [Cause and Effect] alone can we go beyond the evidence of our memory and senses. (EHU 4.4 / 26) 2. All reasonings concerning matter of fact seem to be founded on the relation of Cause and Effect. (EHU 4.4 / 26) . . . all arguments concerning existence are founded on the relation of cause and effect . . . (EHU 4.19 / 35) . . . all our evidence for any matter of fact, which lies beyond the testimony of sense or memory, is derived entirely from the relation of cause and effect . . . (EHU 12.22 / 159) 3. No object ever discovers, by the qualities which appear to the senses, either the causes which produced it, or the effects which will arise from it . . . (EHU 4.6 / 27) 90

9780826443595_Ch03_Final_txt_print.indd 90

2/9/2012 9:38:55 PM

9780826443595_Ch03_Final_txt_print.indd 91

(1) Only the relation of cause and effect can take us beyond the evidence of our memory and senses

(2) All factual inferences to the unobserved are founded on the relation of cause and effect

(4) Any effect is quite distinct from its cause, and many different effects are equally conceivable

(7) All reasonings from experience are founded on the Uniformity Principle (UP)

(6) All factual inferences to the unobserved are founded on experience

(20) CONCLUSION

(8) All factual inferences to the unobserved are founded on UP

(3) Sensory perception of any object does not reveal either its causes or its effects, and there is no known connexion between the sensible qualities and its ‘secret powers’ (9) UP is not founded on anything that we learn through the senses about objects’ ‘secret powers’

(14) A change in the course of nature can be distinctly conceived, and hence is possible

(13) Two kinds of argument are available (for proving UP): demonstrative and factual

(15) Future uniformity cannot be inferred demonstratively from past uniformity

(16) If there is a good argument for UP, it must be a factual inference

(17) Any factual inference to UP would be circular

(18) There is no good argument of any kind for UP

(10) UP can be founded on reason only if it is founded on experience (of uniformity)

(19) UP is not founded on reason (12) UP can be founded on reason only if it is founded on argument (via some medium enabling it to be inferred from past experience of uniformity)

(11) The inference from past uniformity to future uniformity is not intuitive

2/9/2012 9:38:55 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

91

(5) Causal relations cannot be known a priori, but can only be discovered by experience

No factual inference to the unobserved is founded on reason

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

6.

7.

8.

9.

10.

the assistance of observation and experience. (EHU 4.11 / 30) . . . our knowledge of that relation [of cause and effect] is derived entirely from experience . . . (EHU 4.19 / 35) . . . nor can our reason, unassisted by experience, ever draw any inference concerning real existence and matter of fact . . . (EHU 4.6 / 27) In vain, therefore, should we pretend to determine any single event . . . without the assistance of observation and experience. (EHU 4.11 / 30) . . . we always presume, when we see like sensible qualities, that they have like secret powers, and expect, that effects, similar to those which we have experienced, will follow from them . . . (EHU 4.16 / 33) We have said, that . . . all our experimental conclusions proceed upon the supposition, that the future will be conformable to the past . . . (EHU 4.19 / 35) . . . all inferences from experience suppose, as their foundation, that the future will resemble the past, and that similar powers will be conjoined with similar sensible qualities . . . (EHU 4.21 / 37) [This proposition is implicit in the inferential sequence:] We have said, that all arguments concerning existence are founded on the relation of cause and effect; that our knowledge of that relation is derived entirely from experience; and that all our experimental conclusions proceed upon the supposition, that the future will be conformable to the past. (EHU 4.19 / 35) . . . the mind is not led to form such a conclusion concerning [sensible qualities and secret powers’] constant and regular conjunction, by any thing which it knows of their nature . . . (EHU 4.16 / 33) [This proposition is implicit in Hume’s transition from considering ‘a priori’ evidence for the Uniformity Principle

11.

12.

13.

14.

15.

16.

to considering experiential arguments for it:] As to past Experience, it can be allowed to give direct and certain information of those precise objects only, and that precise period of time, which fell under its cognizance: but why this experience should be extended to future times, and to other objects, which for aught we know, may be only in appearance similar; this is the main question on which I would insist. (EHU 4.16 / 33–4) The connexion between these propositions [I have found that such an object has always been attended with such an effect and I foresee, that other objects, which are, in appearance, similar, will be attended with similar effects] is not intuitive. (EHU 4.16 / 34) There is required a medium, which may enable the mind to draw such an inference, if indeed it be drawn by reasoning and argument. (EHU 4.16 / 34) All reasonings may be divided into two kinds, namely demonstrative reasoning, or that concerning relations of ideas, and moral reasoning, or that concerning matter of fact and existence. (EHU 4.18 / 35) . . . it implies no contradiction, that the course of nature may change . . . May I not clearly and distinctly conceive [such a thing]? (EHU 4.18 / 35) That there are no demonstrative arguments in the case, seems evident . . . (EHU 4.18 / 35) . . . whatever is intelligible, and can be distinctly conceived, implies no contradiction, and can never be proved false by any demonstrative argument or abstract reasoning a priori . . . (EHU 4.18 / 35) If we be, therefore, engaged by arguments to put trust in past experience, and make it the standard of our future judgment, these arguments must be probable only, or such as regard matter of fact and real existence . . . (EHU 4.19 / 35)

92

9780826443595_Ch03_Final_txt_print.indd 92

2/9/2012 9:38:56 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION 17. To endeavour, therefore, the proof [that the future will be conformable to the past] by probable arguments, or arguments regarding existence, must be evidently going in a circle, and taking that for granted, which is the very point in question. (EHU 4.19 / 35–6) 18. . . . it may be requisite . . . to shew, that none of [the branches of human knowledge] can afford such an argument . . . (EHU 4.17 / 35) . . . we have no argument to convince us, that objects, which have, in our experience, been frequently conjoined, will likewise, in other instances, be conjoined in the same manner . . . (EHU 12.22 / 159) 19. . . . it is not reasoning which engages us to suppose the past resembling the future, and to expect similar effects from causes, which are, to appearance, similar . . . (EHU 4.23 / 39) . . . nothing leads us to [expect constant conjunctions to continue] but custom or a certain instinct of our nature . . . (EHU 12.22 / 159) 20. I say then, that, even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding. (EHU 4.15 / 32) . . . in all reasonings from experience, there is a step taken by the mind, which is not supported by any argument or process of the understanding. (EHU 5.2 / 41) All belief of matter of fact or real existence [is due merely to] a species of natural instincts, which no reasoning or process of the thought and understanding is able, either to produce, or to prevent. (EHU 5.8 / 46–7)

2

3

4

5

6

7

NOTES 8 1

The argument appears in Treatise 1.3.6 / 86–94, Abstract 8–16 / 649–52, and Enquiry 4 /

25–39, and is outlined in sections 2.1–4 below. In discussions of induction it is commonly referred to as ‘Hume’s famous argument’, a convenient shorthand that I shall adopt. Note also that ‘induction’ is the modern term for the topic of his argument; he himself never uses the word in this sense. This is the summary of the Section 4 argument alluded to earlier. Note, however, that the previous clause brings in a point from the Section 7 discussion of the idea of necessary connexion, which does not feature in Section 4 itself. Hume does not reject the Causal Maxim, but says that it ‘must . . . arise from observation and experience’ (THN 1.3.3.9 / 82), hinting that he will return to it later (though he never does). For detailed discussion, see Peter Millican, ‘Hume’s Determinism’, Canadian Journal of Philosophy 40 (2010), pp. 611–42, sects II, IV, VI. Section 1.3.7 will in due course move on to the third component, ‘the nature and qualities’ of the belief-idea. Hume continues to mention the imagination’s power to mix and separate its ideas (e.g. Abs. 35 / 662, EHU 5.10 / 47–8), but the Separability Principle as such is never again invoked as it had been, very significantly, in the Treatise (e.g. THN 1.1.7.3 / 18–9, 1.2.3.10 / 36–7, 1.2.5.3 / 54–5), arguably sometimes with absurd results (e.g. THN 1.4.3.7 / 222, 1.4.5.5 / 233, 1.4.5.27 / 245–6, App. 12 / 634). At THN 1.3.2.11 / 77, Hume had stressed that (single-case) contiguity and succession are insufficient to characterize a cause and effect relationship, pointing out that ‘There is a necessary connexion to be taken into consideration’. Now at THN 1.3.6.3 / 87, he reminds us that ‘Contiguity and succession are not sufficient to make us pronounce any two objects to be cause and effect’, and he expresses satisfaction at having unexpectedly ‘discover’d a new relation . . . This relation is their constant conjunction.’ The link between the passages is evident both from the content and the capitalization. John Locke, An Essay concerning Human Understanding, ed. Peter H. Nidditch, (Oxford: Clarendon Press, 1975), IV.xv.1, IV.xvii.2. Humean demonstration corresponds to what is now called deductive reasoning, in the informal sense of an argument whose premises

93

9780826443595_Ch03_Final_txt_print.indd 93

2/9/2012 9:38:56 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

9

10

11

conceptually guarantee the truth of the conclusion. For more on this, see Peter Millican, ‘Humes Old and New: Four Fashionable Falsehoods, and One Unfashionable Truth’, Proceedings of the Aristotelian Society, supp. vol. 81 (2007), pp. 163–99, sect. V. This result comes from Hume’s theory of relations, at THN 1.3.2.1–3 / 73–4 (for criticism, see Peter Millican, ‘Hume’s Fork, and His Theory of Relations’, forthcoming in Philosophy and Phenomenological Research). In brief, THN 1.1.5 / 13–5 enumerates what Hume takes to be the seven different kinds of relation, which THN 1.3.1.1 / 69–70 then divides into two classes. The four relations ‘that depend solely on ideas’ are the sources of strict ‘knowledge’, with resemblance, contrariety and degrees in quality amenable to intuition (THN 1.3.1.2 / 70), and proportions in quantity or number the basis for demonstration. Of the three ‘inconstant’ relations, identity and relations of time and place are amenable to perception (THN 1.3.2.2 / 73–4), leaving causation as ‘the only one, that can be trac’d beyond our senses, and informs us of existences and objects, which we do not see or feel’ (THN 1.3.2.3 / 74). Hume thus identifies probable with causal reasoning, and the rest of Book 1, Part 3, entitled ‘Of Knowledge and Probability’, is accordingly devoted to ‘the idea of causation . . . tracing it up to its origin’ (THN 1.3.2.4 / 74–5). Strangely, the word ‘probability’ does not appear at all in this Part before THN 1.3.6.4 / 89, except in the title of the Part itself and of Section 1.3.2: ‘Of Probability; and of the Idea of Cause and Effect’. Notice that causal inference is categorically stated to be founded on that presumption – there is no suggestion here of the conditionality that we had at THN 1.3.6.4 / 89: ‘If reason determin’d us, it wou’d proceed upon that principle . . .’ (emphasis added). Nor is there any such conditionality at THN 1.3.6.11 / 91–2, or in either the Abstract (Abs. 13–14 / 651) or the Enquiry (EHU 4.19 / 35–6, 4.21 / 36–7, 5.2 / 41–2). THN 1.3.6.7 / 90 expresses the circularity in causal terms: ‘The same principle cannot be both the cause and effect of another’, apparently in order to make a joke: ‘and this is, perhaps, the only proposition concerning that

12

13

14

15

16

17

18

19

relation, which is either intuitively or demonstratively certain’. The Abstract and Enquiry make clear that the circularity is logical. Before drawing this conclusion, Hume adds (what I have called) a ‘coda’ to his argument (THN 1.3.6.8–10 / 90–1), dismissing an attempt to get round it by appeal to objects’ powers. This attempt is refuted by the simple observation that induction needs to be presupposed to enable us to draw an inference from the powers of past objects to the powers of future objects. For discussion of this coda and its implications, see Peter Millican, ‘Hume’s Sceptical Doubts concerning Induction’, in Peter Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), pp. 107–73, sects 9–9.2. He also refers back to it in a footnote at THN 1.3.14.17 / 163, feeding into his discussion of the idea of necessary connexion. For discussion of some of the nuances of terminology for referring to this kind of reasoning, see Millican, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 3.1, which distinguishes between probable inference, factual inference, factual inference to the unobserved, and inductive inference. Hume generally takes for granted that all of these coincide. The argument from THN 1.3.6.8–10 / 90–1 is also very briefly summarized, in the last two sentences of paragraph 15. For more on this ‘coda’, see note 12 above. Notice that Hume seems entirely happy to take perception and memory for granted here, fitting with the strategy described in section 1 above, of allowing default authority to our faculties. Scepticism regarding the senses is addressed at THN 1.4.2–4 / 187–231 and EHU 12.6–16 / 151–5, but Hume’s ultimate attitude to it remains far less clear than his position on induction. This notion of a proof plays a significant role in Hume’s argument concerning miracles in Section 10 of the Enquiry. This is taken from Millican, ‘Hume’s Sceptical Doubts concerning Induction’. Thus there is no evidence here, as influentially claimed by David Stove, Probability and Hume’s Inductive Scepticism (Oxford: Oxford University Press, 1973), p. 50, that Hume’s method of argument shows him to be a

94

9780826443595_Ch03_Final_txt_print.indd 94

2/9/2012 9:38:57 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

20

21

22

23

‘deductivist’, presupposing that only deductively valid arguments are legitimate. A similar point, though less obvious, can be made about the Abstract (‘The mind can always conceive any effect to follow from any cause, and indeed any event to follow upon another’, Abs. 11 / 650) and the Treatise (‘When we pass from a present impression to the idea of any object, we might possibly . . . have substituted any other idea in its room’, THN 1.3.6.1 / 87). This case of applied mathematics (cf. also THN 2.3.3.2 / 413–4) shows that Hume is quite comfortable with demonstrative, mathematical reasoning being applied to a posteriori premises. For discussion of this point, see Millican, ‘Humes Old and New’, sect. V. Hume’s talk of ‘secret powers’ is new in the Enquiry, and seems to reflect a more sophisticated understanding of scientific reasoning than is evident in the Treatise and Abstract. In the Treatise, science is generally treated as involving predictions of discrete types of event based on previous patterns of conjunction or difference (as in the ‘rules by which to judge of causes and effects’ of THN 1.3.15 / 173–176). The Enquiry, by contrast, evinces an awareness (e.g. at EHU 4.13 / 31 and EHU 7.29n17 / 77n) that science more typically deals with events having continuously varying characteristics – such as the velocity of a billiard ball – whose prediction involves the interplay of mathematically determined forces. For more on this, see Peter Millican, ‘Against the New Hume’, in Rupert Read and Kenneth A. Richman (eds), The New Hume Debate: Revised Edition (London: Routledge, 2007), pp. 211–52, at pp. 232–3. Hume obviously means us to infer accordingly – though he does not explicitly state – that [8] all factual reasoning, since it has to be founded on experience, presupposes such a resemblance (i.e. the Uniformity Principle). See also note 27 below. In EHU 4.16 / 33–4 itself, Hume oscillates between reference to the activity of inference from observed to unobserved, and to the presupposition of resemblance on which such inference is based. Indeed it seems that he takes the foundation of the inference to be the same as the foundation of the presupposition that it manifests. This supports an interpretation of

24

25

26

27

28

29

the Uniformity Principle as implicit rather than explicit, a principle we exhibit by our inferential behaviour rather than one we always consciously consider. Such an interpretation nicely squares Hume’s repeated commitment to the Principle’s role in all inductive inference (see note 10) with his clear recognition at THN 1.3.8.13 / 103–4 that we characteristically ‘draw inferences from past experience, without reflecting on . . . that principle’. See also note 31 below. This suggests that if the inference were intuitive, it would count as ‘reasoning and argument’ notwithstanding the lack of a ‘medium’. Indeed, as we shall see later (section 3.1), in Hume’s day the terms ‘reasoning’ and ‘argument’ did not imply complex ratiocination. Hume is fond of elegant variation, frequently using a variety of terms for the same concept. ‘Moral reasoning’, ‘reasoning concerning matter of fact and real existence’, ‘probable arguments’ and ‘arguments concerning existence’ are all ways of referring to what we are here calling factual reasoning. See note 14 above. That he takes these to be equivalent was made clear by EHU 4.2 / 25–6, where he first explained the notion of a matter of fact. For an earlier occurrence of this last implicit inference, see note 22 above. As in the Treatise (note 12 above) and Abstract (note 15 above), Hume rounds off the argument in the Enquiry with a coda (EHU 4.21 / 36–8) in which he refutes the attempt to circumvent his argument by appeal to objects’ powers. He also adds a parting shot at EHU 4.23 / 39 which emphasises the unlikelihood that peasants, infants or ‘brute beasts’ should form their inductive expectations on the basis of ‘any process of argument or ratiocination’. Though the point is well made, however, its philosophical significance is less clear, because those who take induction to be rationally founded need not be committed to supposing that animals (etc.) function purely rationally – see Millican, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 9.3. Recall that [12] is the claim that ‘There is required a medium, which may enable the mind to draw such an inference, if indeed it be drawn by reasoning and argument.’

95

9780826443595_Ch03_Final_txt_print.indd 95

2/9/2012 9:38:57 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

30

31

32

33

(EHU 4.16 / 34) – that is, because the Uniformity Principle cannot be established directly through sensory perception or intuition, if it is to be established by reason at all, then this must be on the basis of some stepwise argument or ratiocination. The other implicit final stages are also stated explicitly elsewhere: [18] ‘we have no argument to convince us, that objects, which have, in our experience, been frequently conjoined, will likewise, in other instances, be conjoined in the same manner’ (EHU 12.22 / 159); [19] ‘it is not reasoning which engages us to suppose the past resembling the future, and to expect similar effects from causes, which are, to appearance, similar’ (EHU 4.23 / 39). See note 23 above for the nature of this presupposition, which need not be conscious, but is implicitly manifested by the making of the inference. So UP need not take any very explicit or determinate form (contrary to the impression given by THN 1.3.6.4 / 88–9), and is best understood as something like a general principle of evidential relevance between observed and unobserved, more in line with the expression of the Enquiry: ‘we . . . put trust in past experience, and make it the standard of our future judgment’ (EHU 4.19 / 35); we take ‘the past [as a] rule for the future’ (EHU 4.21 / 38). This seems right: in taking such an inference to be better informed than an a priori inference, we are ipso facto presuming that what happened in the past provides evidence that is positively relevant to what will happen in the future. For more on the Uniformity Principle and its presupposition, see Millican, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 3.2 and especially sect. 10.2. Moreover this sequence of argument seems to be entirely deliberate, because it occurs very explicitly twice, first within the main argument at EHU 4.16 / 32–4, and then again in the coda at EHU 4.21 / 37. In the conclusion of Book 1 of the Treatise, Hume’s attempt to meet the sceptical challenge says very little about the issue of induction, except as part of a general concern regarding the role of ‘the imagination, or the vivacity of our ideas’ (THN 1.4.7.3 / 265). There the more pressing problems are those that threaten inevitable error and contradiction (notably

34

35 36

37

38

39

the existence of body, the metaphysics of causation, and the self-annihilation of reason), which the simple assumption of uniformity never does. The Enquiry’s response to the Pyrrhonian sceptic, starting from a rejection of extreme antecedent scepticism, would not be nearly as effective against varieties of consequent scepticism that bring to light genuine contradictions – rather than simply lack of ultimate grounding – in our faculties, and this might explain why Hume very much downplays these more problematic topics in the Enquiry. His attitude to them seems to be that they are best left alone: for example, metaphysical enquiries into the nature of matter are likely to lead to contradiction or unintelligibility (EHU 12.14–15 / 153–5) unless, perhaps, we fall back on a notion of matter so empty as to be unexceptionable (EHU 12.16 / 155). Hume’s final recommendation is for a mitigated scepticism that inspires a suitable degree of ‘doubt, and caution, and modesty’ (EHU 12.24 / 162), and which also focuses our enquiries on ‘such subjects as are best adapted to the narrow capacity of human understanding’ (EHU 12.25 / 162), notably those where we are able to progress either through mathematical demonstration (EHU 12.27 / 163) or induction from experience (EHU 12.28–31 / 163–6). David Owen, Hume’s Reason (Oxford: Oxford University Press, 1999), p. 136. Ibid., p. 132. Barry Stroud, Hume (London: Routledge & Kegan Paul, 1977), p. 54. Antony Flew, Hume’s Philosophy of Belief (London: Routledge & Kegan Paul, 1961). Stove, Probability and Hume’s Inductive Scepticism. Stroud himself (Hume, pp. 56–7) reacted against this, suggesting that what he saw as Hume’s extreme scepticism could more plausibly be attributed to a ‘potentially regressive aspect of the notion of reason or justification’ whereby evidence E can count as a reason for believing P only if one has some reason R for taking E as a reason. If we then ask about the basis for R in turn, and continue in this way, we get a regress which can apparently be terminated only by ‘something we could not fail to be reasonable in believing’ (ibid.,

96

9780826443595_Ch03_Final_txt_print.indd 96

2/9/2012 9:38:58 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

40 41

42

43

44

45

p. 62), such as an immediate experience or selfevident truth. Hume’s invoking of UP within his argument is indeed somewhat in this spirit, but when considering UP’s own foundation, he seems content to stop with (fallible) sensation or memory, not only with the certainty of intuition or demonstration, while the appeal to factual inference generates a circle rather than an infinite regress. Nevertheless Stroud’s account is illuminating, in stressing the seductive assumption that justification is required at each step if scepticism is to be resisted. Hume’s strategy outlined in section 1 above rejects this by shifting the onus onto the sceptic. Flew, Hume’s Philosophy of Belief, p. 171. For some other passages in a similar spirit, see note 125 below. Tom Beauchamp and Thomas Mappes, ‘Is Hume Really a Sceptic about Induction?’, American Philosophical Quarterly 12 (1975), pp. 119–29. Barbara Winters, ‘Hume on Reason’, Hume Studies 5 (1979), pp. 20–35 was perhaps most influential in promoting the idea that Hume’s notion of reason is ambiguous in this way. Tom Beauchamp and Alexander Rosenberg, Hume and the Problem of Causation (Oxford: Oxford University Press, 1981); N. Scott Arnold, ‘Hume’s Skepticism about Inductive Inference’, Journal of the History of Philosophy 21 (1983), pp. 31–55; Janet Broughton, ‘Hume’s Skepticism about Causal Inferences’, Pacific Philosophical Quarterly 64 (1983), pp. 3–18; and Annette C. Baier, A Progress of Sentiments: Reflections on Hume’s Treatise (Cambridge MA: Harvard University Press, 1991) all gave slightly different antideductivist readings, some of the nuances of which are discussed by Don Garrett, Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997), pp. 83–91. Peter Millican, ‘Hume’s Argument concerning Induction: Structure and Interpretation’, in Stanley Tweyman (ed.), David Hume: Critical Assessments (London: Routledge, 1995), vol. 2, pp. 91–144; Garrett, Cognition and Commitment. Most obviously, the famous argument treats probable argument as a potential foundation for the Uniformity Principle, whereas a

46

47 48 49 50

51

52

53

54

deductivist must consider any merely probable argument as evidentially worthless from the start. If Hume were a deductivist, indeed, then he could dismiss induction in a single step with his Conceivability Principle. For more detail on all this, see Millican, ‘Hume’s Argument concerning Induction’, pp. 123–4, 136; ‘Hume’s Sceptical Doubts concerning Induction’, pp. 155–6; Garrett, Cognition and Commitment, pp. 86–8). See in particular the passages quoted near the end of the first paragraph of section 3 above. These and others are cited in this connexion by Millican, ‘Hume’s Argument concerning Induction’, pp. 127, 136; ‘Hume’s Sceptical Doubts concerning Induction’, pp. 161–2; Garrett, Cognition and Commitment, pp. 85–6. Locke, Essay, IV.xvii.2. e.g. Locke, Essay, IV.xv.2. Garrett, Cognition and Commitment, p. 92. Locke’s usage is somewhat variable, though I consider perception to be more fundamental to Lockean reason than inference (see my ‘Hume’s Argument concerning Induction’, p. 137, or for more detail, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 2). Note that both of these are distinct from the intermediate idea characteristic which Owen considers fundamental to Lockean reason. Garrett, Cognition and Commitment, pp. 91–2. The modified interpretation first appeared in Don Garrett, ‘Ideas, Reason and Skepticism: Replies to My Critics’, Hume Studies 24 (1998), pp. 171–94, but his 2002 piece ‘The Meaning of Hume’s Conclusion concerning “Inductive” Inferences’ (in Peter Millican, Reading Hume on Human Understanding, pp. 332–4) was based directly on the two relevant sections of his 1997 book (Cognition and Commitment), reworded accordingly. Garrett, ‘The Meaning of Hume’s Conclusion’, p. 333. Or, presumably, A to C. Garrett’s suggestion (‘Ideas, Reason and Skepticism’, pp. 182–3) is that in attempting to infer from A to D, we first observe that A implies B intuitively (i.e. self-evidently), leaving a gap between B and D. We then set out to fill that gap, by noticing that B implies C, and C implies D. We put

97

9780826443595_Ch03_Final_txt_print.indd 97

2/9/2012 9:38:58 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

55

56

57

58

59

60

61

62

63

these last two implications together, deducing that B implies D (this is Garrett’s intermediate inference). Now from A implies B and B implies D, we can deduce that A implies D. See, for example, Peter Millican, ‘Hume on Reason and Induction: Epistemology or Cognitive Science?’, Hume Studies 24 (1998), pp. 141–60, at pp. 145–7, and David Owen, ‘A Reply to My Critics’, Hume Studies 26 (2000), pp. 323–37, at pp. 329–30. Samuel Johnson, A Dictionary of the English Language (London, 1756). ‘To deduce’ is defined in three clauses: ‘1. To draw in a regular connected series. 2. To form a regular chain of consequential propositions. 3. To lay down in regular order.’ ‘Ratiocination’ is defined in just one clause: ‘The act of reasoning; the act of deducing consequences from premises.’ Also, of course, Hume’s own theory of inductive inference implies that it typically does not proceed in a stepwise manner, but essentially reduces to conception (see THN 1.3.7.5n20 / 96–7n); yet he never hints that terms such as ‘argument’, ‘inference’, ‘proof’ or ‘reasoning’ are thereby rendered inappropriate to these transitions of thought. So it is hard to see how he could consistently refuse to apply them – on grounds of immediacy – to ‘intuitive inference’. Such an emphasis comes later, with the positive account in terms of instinctive custom (e.g. at THN 1.3.8.13 / 103–4 and EHU 5.8 / 46–7). At least, this looks like a plausible implication, just as one invalid step within a mathematical proof is enough to render the entire proof invalid. But as we shall see later (section 4), things are not quite so straightforward here. For another way of sharpening this sort of objection, see Millican, ‘Humes Old and New, sect. VI, which expands on Millican, ‘Sceptical Doubts concerning Induction’, pp. 158–60. There I focus on the very last step of Hume’s argument, whereby he concludes that because factual inference is founded on UP, and UP is not founded on reason, it follows that factual inference is not founded on reason. This step looks very dubious if ‘reason’ here is supposed to mean stepwise ratiocination (or, indeed, higher-level argument). See note 10 above for equivalent passages in the Abstract and Enquiry versions.

64

65

66

67

See for example Owen, Hume’s Reason, pp. 9–10, 120–2, 127–30, 141, 148. The books of both Owen and Garrett present only the Treatise version of the argument, and indeed Owen’s analysis hardly mentions the Abstract or Enquiry. Garrett takes relevant quotations from the later works, but states without analysis that ‘the structure and language of the other versions of the argument are parallel’ to that in the Treatise (Cognition and Commitment, p. 82). Descartes’s Meditations, for example, presents the Ontological Argument for the existence of a perfect God, and then appeals to God’s non-deceptive nature to vindicate various factual beliefs about the unobserved, all apparently without any essential reference to the Uniformity Principle. Or – if we take the Enquiry version – in any deliverance of the senses or intuition. This objection goes back to my ‘Hume on Reason and Induction’, sect. VII (1998), and is also discussed by Garrett, ‘Ideas, Reason and Skepticism’, p. 187; Millican ‘Hume’s Sceptical Doubts concerning Induction’, pp. 157–8; Louis Loeb, ‘Psychology, Epistemology, and Skepticism in Hume’s Argument about Induction’, Synthese 152 (2006), pp. 321–38, at pp. 328–9; and Abraham Sesshu Roth, ‘Causation’, in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), pp. 95–113, at pp. 108–11. The case of faulty factual arguments (e.g. in Hume’s coda at THN 1.3.6.10 / 91 and EHU 4.21 / 36–8) is less clear, because if they presuppose UP, then the famous argument – as interpreted by Garrett and Owen – can still get a purchase on them. For critical discussion, see Loeb, ‘Psychology, Epistemology, and Skepticism’ (p. 329), who goes on to suggest his own explanation of why Hume fails to consider faulty arguments here: ‘The answer must be that Hume imposes an epistemic constraint on any causal explanation of inductive inference: the explanation of our making inductive inferences must be compatible with their being justified’ (p. 330). Helen Beebee, Hume on Causation (London: Routledge, 2006), pp. 55–6, takes a similar line, and both are discussed in my ‘Humes Old and New’, pp. 186–8. In brief, I find their approach textually unsupported and also in tension with

98

9780826443595_Ch03_Final_txt_print.indd 98

2/9/2012 9:39:00 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

68

69 70

71 72 73 74

75

76 77 78

79

80 81

82

the sceptical tone of Hume’s famous argument and of his later references to it. A far simpler solution is to see ‘reason’ as referring to our cognitive faculty – see section 3.2 below. Richard Price, A Review of the Principal Questions and Difficulties in Morals (London: 1758), p. 34. Ibid., p. 40n. Owen also adds a note at this point: ‘The qualification, “prior to our engaging in probable reasoning”, is important, because Hume thinks that once we are engaged in the practice of probable reasoning, we come to believe the uniformity principle and use it in probable reasoning. . . . This requires an account of how we first engage in probable reasoning, before the principle is available to us.’ (Hume’s Reason, p. 128n30) Owen, Hume’s Reason, pp. 128–30. Garrett, Cognition and Commitment, p. 94. Owen, Hume’s Reason, pp. 132–4. Garrett, ‘Ideas, Reason and Skepticism’, pp. 180–1. Unlike Garrett and Owen, however, I do not take the principle in question, ‘that like objects, plac’d in like circumstances, will always produce like effects’, to be identical to the Uniformity Principle. The former concerns the consistency of events within our experience, whereas UP concerns the evidential relevance of observed to unobserved. Without UP, experienced consistency (or, indeed, any other experienced pattern) could not be extrapolated from past to future. Owen, Hume’s Reason, pp. 131, 170–1. Garrett, ‘Ideas, Reason and Skepticism’, p. 184 Millican, ‘Hume on Reason and Induction’, p. 153. This is the main respect in which my own views have changed over time, largely in response to Don Garrett’s criticisms. Most other aspects of my previous interpretation remain in place; for example it will become clear in section 4 below that a perceptual notion of reason makes a highly plausible Humean target, even if we do not suppose that he was employing such a notion himself within the famous argument. Garrett, Cognition and Commitment, p. 92. Garrett, ‘The Meaning of Hume’s Conclusion’, p. 333.

83

84

85

86

87

88

89 90

Garrett, ‘Ideas, Reason and Skepticism’, p. 184. Ibid. It seems to be a logical rather than psychological point that some such step must be present in every inductive inference, given that – as Garrett acknowledges – the ‘supposition of UP’ that it exhibits can be entirely unconscious. Locke starts his chapter ‘Of Reason’ (Essay, IV.xvii) with the remark that ‘The Word Reason in the English Language has different Significations: sometimes it is taken for true, and clear Principles: Sometimes for clear, and fair deductions from those Principles: and sometimes for the Cause, and particularly the final Cause.’ He then goes on to say that his chapter concerns yet another signification, for that ‘Faculty in Man . . . whereby Man is supposed to be distinguished from Beasts, and wherein it is evident he much surpasses them’. Earlier, at Essay II.xxi.17–20, Locke ridicules the language of ‘faculties’ as a source of philosophical error. For more on his view, see Millican, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 2, and cf. note 50 above. As acknowledged in Hume’s letter of 10 January 1743 (LDH 1.45, 19). Francis Hutcheson, Synopsis of Metaphysics (1744), trans. Michael Silverthorne, in Francis Hutcheson, Logic, Metaphysics, and the Natural Sociability of Mankind (Indianapolis: Liberty Fund, 2006) Francis Hutcheson, A Short Introduction to Moral Philosophy (1747), ed. Luigi Turco (Indianapolis: Liberty Fund, 2007), p. 25. Francis Hutcheson, An Essay on the Nature and Conduct of the Affections. With Illustrations on the Moral Sense, 3rd edn (London: 1742), pp. 30–1. Ibid., pp. 219–20. Note that the quoted paragraph also treats ‘the Intellect’ as just another elegant variation on ‘Reason’ and ‘the Understanding’. Hume does the same, albeit more rarely (DNR 3.153, 3.156), though he quite often refers in a similar spirit to the ‘intellectual faculties’ (THN 1.3.12.20 / 138, 2.3.8.13 / 437; EHU 5.5n8 / 43–4n, 9.6 / 108; EPM 1.9 / 173, EPM App. 1.11 / 290, 13 / 291, 18 / 293, 3.9 / 307; ‘Of the Standard of Taste’, E 240–1). Garrett talks of Hume as giving ‘an argument against the intellect’ (Cognition and Commitment, p. 20), but this is misleading unless ‘the intellect’ here

99

9780826443595_Ch03_Final_txt_print.indd 99

2/9/2012 9:39:00 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

91

92

93

94

is understood to mean ‘the intellect conceived of as a faculty of non-sensory ideas’ (a conception that Garrett traces through Descartes, Spinoza and Leibniz, but is not shared by Locke, Berkeley or Hume). See, for example, Anthony Ashley Cooper, Lord Shaftesbury, Characteristicks of Men, Manners, Opinions, Times, 3 vols, 3rd edn [the edition purchased by Hume in 1726] (London, 1723), vol. 2, II.ii, p. 118; Joseph Butler, The Analogy of Religion, Natural and Revealed, to the Constitution and Course of Nature, 2nd edn, corrected (London, 1736), I.vi, p. 174; Price, Review, I ii, p. 23. See, for example, Henry Home, Lord Kames, Essays on the Principles of Morality and Natural Religion, 2nd edn (London, 1758), essay VII, p. 268n, and James Beattie, An Essay on the Nature and Immutability of Truth; in Opposition to Sophistry and Scepticism (Edinburgh, 1770), I.i, pp. 37–8. James Oswald, An Appeal to Common Sense in Behalf of Religion (Edinburgh, 1766), I.ii.1, p. 80 and Thomas Reid, Essays on the Intellectual Powers of Man (Edinburgh, 1785), VII.i, p. 671 are likewise keen to insist on a narrow use of ‘reasoning’, distinguished from ‘judging’, though they allow both operations to be subsumed under ‘reason’. For other relevant passages from the Treatise, see, for example, THN 1.3.6.4 / 88–9, 1.3.13.12 / 149–50, 1.4.1.1 / 180, 1.4.1.12 / 186–7, 1.4.2.14 / 193, 1.4.7.7 / 267–8, 2.3.3.2–6 / 413–6, 3.1.1.16–18 / 462–3, and 3.1.1.26 / 468–9. For passages from the Abstract, see Abs. 11 / 650–1, 27 / 657, and from the Enquiry, EHU 4.0–1 / 25, 5.5 / 43, 5.22 / 55, 7.28 / 76 and 9.0–1 / 104. Further examples may be found in Hume’s other works. Together with Hume himself (EHU 1.14 / 13–14, 8.22 / 93), see, for example, René Descartes, Fourth Meditation, in The Philosophical Writings of Descartes, trans. John Cottingham, Robert Stoothoff and Dugald Murdoch (Cambridge: Cambridge University Press, 1984), vol. 2, pp. 39–40; Locke, Essay, II.vi.2; George Berkeley, A Treatise concerning the Principles of Human Knowledge, ed. Jonathan Dancy (Oxford: Oxford University Press, 1998), I.27. References to Hume’s contemporaries Hutcheson, Price and Reid will follow.

95 96

97

98

99

100

101

102 103

Reid, Intellectual Powers, I.vii, pp. 67–8. Note, for example, the general division within the Treatise between Book 1 ‘Of the Understanding’ and Book 2 ‘Of the Passions’ (including Part 3 ‘Of the will and direct passions’), and also the footnote at THN 1.3.7.5n20 / 96, where Hume criticizes the ‘universally receiv’d’ threefold ‘division of the acts of understanding’ which Reid describes. Hutcheson, A Short Introduction to Moral Philosophy, pp. 25–6. Price deprecates ‘the division which has been made by some writers, of all the powers of the soul into understanding and will; the former comprehending under it, all the powers of external and internal sensation, as well as those of judging and reasoning’. By contrast, he says, ‘I all along speak of the understanding, in the most confined and proper sense of it . . . and distinguished from the powers of sensation’ (Review, I.ii, p. 20n). Note, however, that Price implicitly equates the understanding with reason (ibid., p. 23) thus using ‘reason’ in a broader sense than those such as Kames (cf. note 92 above) who exclude intuition from its scope. Hutcheson, Synopsis of Metaphysics, p. 112. The original Latin of the final clause is ‘Ad Intellectum, referentur Sensus’. Sect. 6 ends with a short paragraph on ‘Natural associations of ideas’, which Hutcheson sees as playing an important role in both imagination and memory; sect. 7 briefly discusses what is pleasing or distressing to the senses, and our consequent judgements (of good and evil) and passions; then sect. 8 discusses habit, and sect. 9 relative ideas and judgements. Hutcheson, Illustrations on the Moral Sense, I, p. 215. Price, Review, pp. 19–20. For example Locke: ‘the understanding . . . is the most elevated Faculty of the Soul . . . Its searches after Truth, are a sort of . . . Hunting’ (Essay, Epistle to the Reader, paragraph 1); David Hartley: ‘The Understanding is that Faculty, by which we . . . pursue Truth, and assent to, or dissent from, Propositions.’ (Observations on Man, His Frame, His Duty, and His Expectations, Bath and London, 1749, vol. 1, Introduction, p. iii).

100

9780826443595_Ch03_Final_txt_print.indd 100

2/9/2012 9:39:01 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION 104

105

106

107

108

109

Other passages that identify reason with the discovery of truth, though usually less explicitly, are at THN 2.3.3.3 / 414, 2.3.3.5–6 / 415–6, 2.3.3.8 / 417, 3.1.1.4 / 456–7, 3.1.1.19n69 / 464n, 3.1.1.25–7 / 467–70, 3.2.2.20 / 496, THN App. 1 / 623; EPM 1.7 / 172, EPM App. 1.6 / 287. All of these occur in a context where Hume is contrasting reason (or the understanding) with conative rather than cognitive notions, thus corroborating its identification as the overarching cognitive faculty. This note (which can be found as a Textual Variant on p. 177 of my edition of Hume’s Enquiry concerning Human Understanding, Oxford World’s Classics, 2007) is of particular interest because it credits Hutcheson with distinguishing between ‘the Understanding’ and ‘That Faculty . . . by which we perceive Vice and Virtue’, although Hutcheson himself considered the moral sense to be one of the ‘reflexive or subsequent sensations’, thus falling within the domain of the understanding. Price, Review, p. 12n, mentions Hume’s note in the course of criticizing Hutcheson. Notice that Hume is implying that if we could observe a conjunction of cause and effect, then we could form such a ‘conclusion . . . concerning . . . existence’ and ‘satisfy our reason in this particular’. So as in the famous argument concerning morals (cf. the quotation from THN 3.1.1.12 / 459 near the beginning of section 3.1 above), Hume is clearly here treating causal, factual inference as an operation of reason. Such nuances can apply with many words that are associated with some achievement. For example, a cure that does not work is strictly a contradiction in terms, but it is fairly natural to say, in appropriate circumstances, ‘that cure is useless and ought to be banned’. An analogy here would be to an accounting error within a company, which on a broad interpretation could refer to any error in the accounts (including faulty data from external sources), but on a narrower interpretation would mean an error due to the accountants themselves. On the conative side, Hume hardly ever speaks of ‘faculties’, explicitly referring to ‘the will’ and ‘the passions’ as faculties just once

110

111

112

each (at THN 2.3.3.9 / 417–8 and 2.2.2.16 / 339 respectively). Judgemental ‘taste’ is called a faculty in ‘Of the Standard of Taste’, E 240–1, and spoken of as having ‘a productive faculty’ in a famous passage at EPM App. 1.21 / 294. The model of a canvas is obviously most appropriate to visual ideas, which indeed seem to dominate Hume’s thought, although ideas may correspond to any of the senses – including internal ‘reflection’ – and only the ideas of sight and touch will be spatially arranged (not necessarily within a single space). Note that he takes all of our ideas to be sense-copied; hence as Garrett observes (cf. note 90 above), Hume denies any separate faculty that can take a ‘pure and intellectual view’ of ‘refin’d and spiritual’ ideas, unsullied by sensory input (THN 1.3.1.7 / 72). At least, this seems to be what Hume is saying in THN 1.4.7.3 / 265. At THN 1.3.9.3 / 107–8, he appears instead to take the force and vivacity of the ‘impressions or ideas of the memory’ – like that of ‘an immediate impression’ – as itself constituting assent, thus providing a basis for explaining the assent that derives from causal inference. The relationship between memory and the imagination remains somewhat obscure, though Hume’s talk of ‘impressions of the memory’ (‘somewhat intermediate betwixt an impression and an idea’, THN 1.1.3.1 / 8) suggests that the memory is furnishing ideas that are sufficiently firm and vivid – sufficiently impression-like – to establish copy-ideas in the imagination: ‘The impressions of the memory never change in any considerable degree; and each impression draws along with it a precise idea, which takes its place in the imagination, as something solid and real, certain and invariable.’ (THN 1.3.9.7 / 110). If this is right, then all the ideas that are actually involved in thinking lie within the imagination, and the role of the senses and memory is to supply the ‘impressions’ from which those ideas derive. See THN 1.3.7.6 / 97, 1.3.9.3 / 107–8, 1.3.11.11 / 128, 1.3.14.1 / 155–6, 1.3.14.31 / 169–70; and for references from the Abstract and Enquiry see note 129 below.

101

9780826443595_Ch03_Final_txt_print.indd 101

2/9/2012 9:39:02 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION 113

114

115

116

117

118

119

120 121

I call it a ‘virtual’ stage to reflect Hume’s comment at THN 1.4.6.4 / 253 that ‘The comparison of the theatre must not mislead us. They are the successive perceptions only, that constitute the mind’. THN 1.4.2.22 / 198 gives a case intermediate between passivity and activity, in which the imagination, having been ‘set into any train of thinking, is apt to continue even when its object fails it, and, like a galley put in motion by the oars, carries on its course without any new impulse’. Note that the listed references involving fictions are confined to those that involve characteristic Humean fictions of philosophical interest, rather than arbitrary combinations of ideas (i.e. ‘mere fictions of the imagination’ as at THN 1.3.5.4–5 / 85, 1.3.7.7 / 97, 1.3.9.3 / 108, 1.4.3.1 / 219; EHU 5.12–13 / 49–50, 6.3 / 57). Recall from section 3.2 that the replaced footnote at THN 2.2.7.6n / 371n said ‘the understanding’ here instead of ‘reason’. It is striking, for example, that these are the only two occurrences in Hume’s writings of the phrase ‘offspring of the imagination’. Presumably he was forced to place the footnote at the end of the section to minimize type resetting. The understanding is also identified with the imagination at THN 1.3.8.13 / 104 and 2.3.9.10 / 440. Don Garrett, Cognition and Commitment, p. 92, was, I believe, the first to note this possibility, which is crucial if Hume’s use of ‘reason’ within his argument is to be understood as sincere rather than a target. But of course Garrett’s interpretation of ‘determin’d by reason’ is very different from my own. See, for example, EHU 1.13–14 / 13–14 for the equation of ‘faculties’ with ‘powers’, and also THN Intro. 4 / xv, 1.3.10.9 / 123. The same equation is repeatedly found in Locke, e.g. Essay II.vi.2, II xxi 15, 17, 20. Locke, Essay, II.xxi.20. In a recent debate (Don Garrett and Peter Millican, Reason, Induction and Causation in Hume’s Philosophy, IASH Occasional Paper 19, Edinburgh: Institute for Advanced Studies in the Humanities, University of Edinburgh, 2011), Don Garrett argues that in inductive

122

123

124

125

126

inference, ‘the imagination in the narrow sense is performing the customary transition’ (p. 22), thus denying that custom is reassigned to reason along with ‘our demonstrative and probable reasonings’. Hence he sees the distinction alluded to at THN 1.3.9.19n22 / 117–18 as quite distinct from that drawn at THN 1.4.4.1 / 225, a position I find rather implausible, given that their stated rationale is so similar, namely, to distinguish within the imagination between the principles that ‘are the foundation of all our thoughts and actions’ and those that give rise to ‘whimsies and prejudices’. Even in the Treatise, Hume never says in so many words that custom is a process of the imagination, though THN 1.3.6.4 / 88–9 and 1.3.7.6 / 97 strongly suggest this. In the Enquiry, unlike the Treatise (e.g. THN 1.3.6.16 / 94, 1.3.8.6 / 100–1), Hume is careful to distinguish between the associational relation of causation (discussed at EHU 5.18–19 / 53) and custom (EHU 5.20 / 53–4). We would not usually describe the climber as ‘supported only by’ a rock to which the rope is attached unless the rock was considered potentially less secure than the rope (e.g. suppose the attachment is to a spur of rock that is in imminent danger of cracking – we might well then say that the climber is ‘supported only by the spur’). Together with those quoted earlier, relevant passages include: ‘. . . our reason . . . or, more properly speaking, . . . those conclusions we form from cause and effect . . .’ (THN 1.4.4.15 / 231); ‘. . . these emotions extend themselves to the causes and effects of that object, as they are pointed out to us by reason and experience . . .’ (2.3.3.3 / 414); ‘. . . the operations of human understanding . . . [include] the inferring of matter of fact . . .’ (3.1.1.18 / 463). Both Barry Stroud (Hume, pp. 60–2) and John Kenyon (‘Doubts about the Concept of Reason’, Proceedings of the Aristotelian Society, supp. vol. 59, 1985, pp. 249–67, at pp. 255–7) attribute this to Hume in the context of his argument concerning induction, but neither justifies the attribution, and there is little evidence of it in Hume’s text (cf. note 39 above).

102

9780826443595_Ch03_Final_txt_print.indd 102

2/9/2012 9:39:02 PM

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION 127

128

129

This assumes the internalist perspective which dominated the early modern period. A modern Humean might well take an externalist approach, but given Hume’s explicit response to the sceptic in Enquiry 12 (as described in section 1 above), I am not persuaded by Louis Loeb’s claim that ‘In light of the massive evidence that Hume is not a skeptic about induction, he must reject [the] internalist way of thinking.’ (‘Psychology, Epistemology, and Skepticism’, p. 333). Kenneth Winkler emphasizes how Enquiry 12 supports a more sceptical reading of Enquiry 4: see ‘Hume’s Inductive Skepticism’, in Margaret Atherton (ed.), The Empiricists (Lanham, MD: Rowman & Littlefield, 1999), pp. 183–212, at pp. 193–200. At THN 1.3.9.19 / 117, 1.3.13.9 / 147, 1.3.14.21 / 165, 1.4.4.1 / 225, and 1.4.7.3 / 265. At Abs. 15 / 652, 16 / 652, 21 / 654 (twice), 25 / 656; EHU 5.4–5 / 42–3, 5.5 / 43 (twice), 5.8 / 46, 5.11 / 48, 5.20 / 54, 5.21 / 54–5, 6.4 / 58, 7.29 / 76–7, 8.5 / 82, 8.21 / 92, 9.5 / 106, 9.5n20 / 107, and 12.22 / 159. Note that this contrast cannot be accounted for in terms of Hume’s moving away from the foundational metaphor more generally. On the contrary, he says that induction is ‘founded’ on the relation of cause and effect, or experience, or the Uniformity Principle, and that it is not ‘founded’ on reasoning, argument, or

130

131

132

any process of the understanding, significantly more in the Enquiry – EHU 4.4 / 26, 4.14 / 32 (twice), 4.15 / 32, 4.21 / 37, 9.5 / 106 and 12.29 / 164) – than he does in the Treatise – THN 1.3.6.4 / 88 and 1.3.6.7 / 89–90 (twice). Other aspects of the logic of Hume’s foundational relation are explored in Millican, ‘Hume’s Sceptical Doubts concerning Induction’, sect. 10.1. There is a thematic parallel here with Hume’s account of causation, which is also commonly thought of as sceptical, but in fact provides him with a positive basis for applying causal explanation to the human world. For more on this, see Peter Millican, ‘Hume, Causal Realism, and Causal Science’, Mind 118 (2009), pp. 647–712. For numerous discussions on the topics of reason and induction, I am extremely grateful to Louis Loeb, David Owen and especially Don Garrett, as well as many other members of the Hume Society at its various conferences. I am also grateful to Henry Merivale, Hsueh Qu and especially Dan O’Brien for comments on this paper, and to the Edinburgh Illumni for providing the David Hume Fellowship at Edinburgh, thus giving me the opportunity to work in the delightful context of the Institute for Advanced Studies in the Humanities, overlapping with Don Garrett’s tenure there.

103

9780826443595_Ch03_Final_txt_print.indd 103

2/9/2012 9:39:02 PM

4 THE PSYCHOLOGY AND EPISTEMOLOGY OF HUME’S ACCOUNT OF PROBABLE REASONING Lorne Falkenstein

Over the course of Treatise 1.3, Hume presented what he called a ‘system’ of probable reasoning. He then went on, in Treatise 1.4, to argue that sceptical objections would leave us entirely incapable of belief were our natural inclinations not too strong for philosophical conclusions to be able to restrain our inferences. In the Enquiry concerning Human Understanding he reversed this procedure, first offering ‘sceptical doubts’ about the legitimacy of our inferences concerning matters of fact, then a ‘sceptical solution of these doubts’, and finally a conclusion that we can only be legitimately sceptical of claims in religion and school metaphysics, not everyday experience or natural science. Despite the more optimistic tone, the theory of the Enquiry is built on the same two principles as the ‘system’ of the Treatise: the principle of the association of ideas, and the principle of the genesis of belief in the unobserved as a consequence of association with sensory experience or memories. The common ‘system’ of the Treatise and the Enquiry is sceptical because it takes our beliefs to be the product of naturally occurring psychological

mechanisms rather than logically sound judgement, and declares those beliefs to be ultimately unjustifiable. Despite this sceptical result Hume was able to provide for a logic of probable reasoning, grounded on natural, but unjustifiable beliefs. How he did so is still not well understood. That he was able to do so is one of his great achievements.

1. THE SYSTEM OF THE TREATISE All kinds of reasoning consist in nothing but a comparison, and a discovery of those relations, either constant or inconstant, which two or more objects bear to each other. (THN 1.3.2.2 / 73) In the Treatise Hume’s presentation and defence of his ‘system’ took the form of a torturous journey down the dead-end lanes and twisted turns of a meditative path of discovery, supplemented by appeals to observations and experiments, worries over contrary evidence, and the introduction of refinements to accommodate recalcitrant data. His project was to inquire into the basis of reasoning,

104

9780826443595_Ch04_Final_txt_print.indd 104

2/9/2012 9:39:12 PM

HUME’S ACCOUNT OF PROBABLE REASONING particularly probable reasoning. Reasoning consists in inferring unknown from known objects by means of a relation between the two (THN 1.3.1.2 / 70, 1.3.2.2 / 73). The relations on which all reasoning is based are ‘philosophical relations’, which are discovered by comparing objects with one another (THN 1.3.2.2 / 73). There are also ‘natural relations’. A natural relation is not discovered by comparison or appealed to in order to discover or justify a conclusion. Whether we are aware of it or not, it exercises an influence on the imagination, impelling us to form an idea of an object. This produces a kind of instinctive, counterfeit reasoning (THN 1.1.5 / 13–15, cf. 1.1.4 / 10–13). Relations can be divided into two main kinds: ‘inconstant’ relations, which can alter even while the compared impressions or ideas remain the same (e.g. relations of contiguity and distance in space or time); and ‘constant’ relations, which cannot change without a change in the compared objects (e.g. relations of resemblance). Demonstrative reasoning, yielding certainty, is founded on the latter, while probable reasoning is founded on the former (THN 1.3.1.1–2 / 69–70, 1.3.2.1 / 73). Though we discover a number of constant and inconstant relations by comparing objects, Hume maintained that there is only one relation that can serve as a basis for demonstrative reasoning, the relation of degrees in quantity, and only one that can serve as a basis for probable reasoning, the relation of cause and effect. The other constant relations can always be ‘intuited’ without the need for any demonstration and the other inconstant relations can only be ‘perceived’ and not used as a basis for inference to the unobserved (THN 1.3.1.2 / 70, 1.3.2.2 / 73–4). Thus all probable reasoning reduces to causal reasoning.

2. THE PATH TO THE FIRST PRINCIPLE One of the infelicities of Hume’s presentation is this precipitate assertion that the causal relation is an inconstant relation, and the only relation on which probable reasoning can rest. While Hume never adequately justified the latter claim,1 he went on to give an argument for the former. Our ideas of cause and effect could not be ideas of any of the observed qualities of objects because all objects are causes and effects, and there is no quality that all objects share in common. For any quality we might pick on, there is some object that is a cause or an effect even though it does not have that quality. The idea must therefore be the idea of a relation (THN 1.3.2.5 / 75) – indeed, of a relation that can change while the objects remain the same. At this point a further infelicity arises. Hume simply assumed that causal relations are not immediately perceived or intuited upon comparing objects, as we immediately perceive relations of contiguity or immediately intuit relations of resemblance, but are instead ‘deriv’d’ from some other relation (THN 1.3.2.6 / 75). It is only much later that a reason is given for this assumption (THN 1.3.6.1 / 86–7, explained more fully at Abs. 11–12 / 650–1). If we could perceive or intuit the existence of causal relations, then we would be able to tell upon first acquaintance with a pair of objects whether or not they are causally related. But we cannot do so. Granting that causal relations are not immediately perceived but are instead derived from some other relations, what might those relations be? The only relations we discover when we compare those objects we consider to be causes and effects are contiguity in space and succession in time. But while these relations may hold between causes and effects,2 they also hold between objects

105

9780826443595_Ch04_Final_txt_print.indd 105

2/9/2012 9:39:13 PM

HUME’S ACCOUNT OF PROBABLE REASONING that we consider to be only accidentally conjoined. And, Hume claimed, we think that causes and effects are not just accidentally conjoined but necessarily connected. This is tricky. How can we think that there is a necessary connection between causes and effects if we cannot discover any other relations between them than contiguity in space and succession in time? Occasional suggestions to the contrary notwithstanding,3 Hume did not want to say that we have no idea of necessary connection. A necessary connection is simply a connection that has to be present and cannot be broken. A harness is a connection between a horse and a carriage. If the harness had to run from the horse to the carriage and nothing could break it, it would be a necessary connection. The idea of a necessary connection between causes and effects is similarly the idea of a ‘tye or connexion . . . which binds them together’ (EHU 4.10 / 29). Hume’s claim at this stage in his meditations was not that we are not thinking of anything when we think of a tie reaching across time and space to bind cause and effect together. It was rather that we cannot discover exactly what does the job. We only ever see the horse and the carriage. The apparatus harnessing the two together is not apparent. This does pose a problem: if we cannot discover any harness, why do we think it is there – indeed, that it must be there and cannot be broken? Hume proceeded to try the patience of his reader further by pretending to have no answer to this question and affecting the need to look for one by investigating two related questions: (i) why we consider it necessary that every event have a cause and (ii) why we consider this particular cause necessarily to have that particular effect. In response to the first, he argued that our belief that every event must have a cause could not be

founded on intuition or demonstration and concluded that it must therefore be based on experience (THN 1.3.3 / 78–82). Oddly, he made no attempt to argue that the same answer must be given to the second question. It is only later (THN 1.3.6.1 / 86–7) that a reason is given for concluding that our inferences from cause and effect are not based on demonstration. The reason goes back to an observation on the nature of imagination Hume made at THN 1.1.3.4 / 10. The only limitation on the imagination is that its ideas come from things that have been previously encountered in sensory experience. Once given ideas, it can separate and rearrange them in any way whatsoever. Given an object at a place and a time, the imagination can conceive any object whatsoever at any of the contiguous places at the earlier or later times. Since causes and effects exist at distinct times, any particular cause could be imagined to be followed by anything and any particular effect preceded by anything.4 But were it intuitively obvious or demonstratively provable that this particular cause must have that particular effect, any other alternative would be inconceivable (THN 1.3.6.1 / 86–7).5 Postponing this argument for the moment, Hume (at THN 1.3.4 / 82–4) simply took it for granted that the connection between particular causes and their effects can only be known by experience and proceeded to ask what sort of experience does this. In the process, he presented himself as suddenly discovering a new, third relation obtaining in cases of causality (THN 1.3.6.2–3 / 87–8). The relation is not discovered when comparing individual instances of cause and effect, but only when comparing multiple instances. All instances resembling the cause are spatially contiguous with instances resembling the effect, and precede them in time.

106

9780826443595_Ch04_Final_txt_print.indd 106

2/9/2012 9:39:13 PM

HUME’S ACCOUNT OF PROBABLE REASONING Hume continued to express puzzlement over how this new relation of ‘constant conjunction’ could lead us to conclude that this particular cause must necessarily have that particular effect. We do not discover anything in multiple instances that could not be found in just one instance, and we do not discover anything in one instance that would justify the conclusion. Nor could we take the experience of a constant conjunction to establish the likelihood of a necessary connection, or the likelihood that the causes contain some quality, unknown to us, that gives them the power to bring about their effects.6 Since all we perceive are the observable relations between causes and effects, none of which is a necessary connection, and the observable qualities of the causes, none of which is a power, the most we could infer is that, in the past, objects like the cause have been contiguous to and followed by objects like the effect, and that, in the past, the set of qualities characteristic of the causes has included some further, unknown power. But we are in no position to infer that similar relations must obtain in the future or that similar collections of observed qualities will be accompanied by similar powers in the future (THN 1.3.6.8 / 90 and 1.3.6.10 / 91). The new relation of constant conjunction could only lead us to draw these inferences with the aid of a further supposition, that what has been observed to occur regularly in the past will continue to be observed to occur in the future. But this principle is not demonstrably true, because there is no contradiction in conceiving a change in the course of nature (THN 1.3.6.5 / 89). Moreover, it cannot be proven by appeal to past experience, because the question at issue is precisely why we should take regularities in past experience to establish a rule for what will happen in the future (THN 1.3.6.7 / 89–90).7

At this point in the course of his meditations, Hume finally felt prepared to reveal8 the first principle of the ‘system’. Though constant conjunction provides us with no justification for inferring that causes and effects are necessarily connected, it is a ‘natural’ relation, which impels the imagination to call up an associated idea when presented with its partner. It therefore produces a kind of counterfeit, instinctive ‘reasoning’.9 Observing objects of one sort being customarily followed by objects of another sort trains a habit of thinking into the mind. Once developed, the habit induces the imagination to form an idea of an object upon encountering its customarily conjoined partner, even in the absence of perceiving any tie binding the two together, even in the absence of having any reason to suppose that the future will be like the past, and even in the absence of any recollection of or reflection upon the past instances (THN 1.3.8.13 / 103–4). Although Hume did not draw the conclusion until very much later (THN 1.3.14), this is why, even though we cannot discover any tie that necessarily binds cause and effect together, we think that there must be such a thing.

3. THE PATH TO THE SECOND PRINCIPLE After this long journey of discovery, the second principle of the ‘system’ was quickly, though not easily, uncovered. Hume began by noting that causal inference takes place only when one of the two associated objects is experienced or remembered and the other is not. When both objects are experienced or remembered, there is no occasion to imagine either. And when neither is either experienced

107

9780826443595_Ch04_Final_txt_print.indd 107

2/9/2012 9:39:13 PM

HUME’S ACCOUNT OF PROBABLE REASONING or remembered, we feel impelled to imagine the one upon having occasion to imagine the other, but do not form any belief in the existence of either (THN 1.3.2.2 / 73–4, 1.3.6.2 / 87, 1.3.4 / 82–4). Seeking for an explanation for this variation, Hume noted that the objects of experience and memory are believed to exist (THN 1.3.5.7 / 86). This suggests that the belief we get as a consequence of causal association might be due to some sort of transfer from an experience or memory to an associated object. When an object is believed to exist or have existed, the relation of constant conjunction induces us not merely to conceive an associated object but to form a belief in that object’s existence at a contiguous place and the appropriately earlier or later time (THN 1.3.7.6 / 97).10 Hume was discontented with this bald hypothesis and sought for a justification. At Treatise 1.3.8.2 / 98–9 he attempted to account for the origin of belief as a specific instance of something that can be observed to happen more generally: a natural tendency to confuse readily associated objects. Hume claimed that because the natural transition of thought between objects that have been constantly conjoined is ‘so easy’, it goes unnoticed. Consequently, any mental ‘disposition’ that might happen to attend the latter becomes attached to its impostor. Where the partner is experienced or remembered, the dispositions accompanying experience or memory are confused with the associated idea. Since those dispositions are always bound up with a belief in the existence of the experienced or remembered object, we end up believing in the existence of the associated object as well. This account of the origin of belief is complemented with an account of the nature of belief, infelicitously inserted at Treatise 1.3.4–5 / 82–6, where it interrupts the thread of argument for the first principle, and

only completed at 1.3.7 / 94–8, after having been itself interrupted by 1.3.6 / 86–94. The account begins with an examination of the ‘impressions of the senses and memory’, which are the apparent source of the belief based on causal inference. The relevant point about sense experience had already been made much earlier in Treatise 1.1 / 1–25. Any object that can be sensed can be imagined, so that the difference between sense experience and imagination cannot arise from what is sensed or imagined. Since there is nonetheless a difference, it must be due to something else. Hume referred to this other factor as a different ‘manner’ in which the object is conceived (THN 1.3.7.5 / 96), and tried to describe further this manner of conception by saying that sensing is more ‘forceful’ and ‘vivacious’, imagining ‘fainter’ and ‘lower’ (THN 1.1.1.1 / 1–2, 1.1.1.3 / 2–3). Turning to memory, Hume observed that anything that can be remembered can likewise be merely imagined, and concluded that the difference between what is received as a fantasy and a memory must consist just in the way it ‘feels’ to remember (THN 1.3.5.3–5 / 85–6). Having isolated these differences between sensing and remembering, on the one hand, and imagining on the other, Hume inferred that ‘the belief or assent, which always attends the memory and senses is nothing but the vivacity of those perceptions they present; and that alone distinguishes them from the imagination. To believe is in this case to feel an immediate impression of the senses, or a repetition of that impression in the memory’ (THN 1.3.5.7 / 86). These reflections on the nature of sense and memory led Hume to his conclusion. If the belief attending causal inference arises from a transfer from a sensed or remembered object to an imagined object, and if the belief in sensed or remembered objects is nothing

108

9780826443595_Ch04_Final_txt_print.indd 108

2/9/2012 9:39:13 PM

HUME’S ACCOUNT OF PROBABLE REASONING but a more vivacious conception of those objects, then the belief attending causal inference must likewise be just a more vivacious conception of an object. This is surprising. Rather than find belief in an unperceived object to be the product of a judgement, justified by appeal to a causal relation to an experienced or remembered object, Hume found it to be no different from the belief that attends sense experience and memory. It consists just in a more vivacious conception of the object. Perhaps because he sensed that this conclusion would not be readily accepted, Hume pretended to remain hesitant about it, offering two reasons for his hesitancy: 1. Accounting for the origin of belief involves identifying a cause. According to Hume’s own account of causality, we cannot discover causes merely by inspecting their effects, nor can we be confident that we have identified causes if we have examined only one instance. Either we must find some analogy between the one instance we have before us and other instances and discover some regularity in the succession of events in the analogous cases, or we must show how some combination of more basic, previously established causes could account for the effect. The appeal made at Treatise 1.3.8.2 / 98–9 to a general tendency to confuse readily associated objects is a justification of the latter sort. But Hume went on to declare that, while he would be satisfied if his reader found this reason compelling, he himself placed his chief confidence in being able to uncover a justification of the former sort (THN 1.3.8.3 / 99; App. 3 / 624–5). He wanted to find analogies between the formation of belief as a consequence of experiencing or remembering an object that has been constantly conjoined with some other object in the past and other operations of the mind – something that would allow us

to understand belief as a specific case of something that happens more generally. His search for these analogies had mixed results. They will be discussed later. 2. A more pressing problem is the characterization of belief. Four different characterizations of belief have emerged from what has been said about Hume’s account. At Treatise 1.3.8.2 / 98–9, Hume described belief as a ‘disposition’ of the mind. Over the course of Treatise 1.3.7.7 / 628–9 (from THN App. 1) and 1.3.8.2 / 98–9 this disposition is further described as having to do with drawing and focusing attention (‘rendering more present’, ‘weighing more in the thought’, ‘having more force and influence’, ‘appearing of greater importance’, ‘fixing the attention’); arousing passion (‘elevating the spirits’); inspiring deliberation (‘having a superior influence on the imagination’); and inclining us to action. But THN 1.3.7.4–6 / 95–7 and 1.3.7.7 / 628–9 also describe belief as a ‘manner’ in which an object is conceived. A further characterization of belief is found at THN App. 3 / 624–5 and in passages from the Appendix recommended by Hume for insertion in the main body of the Treatise (1.3.5.5 / 628, 1.3.7.7 / 628–9), where belief is described as a ‘feeling different from the simple conception’ of an object. Finally, and most notoriously, belief is described on numerous occasions in both the Treatise and the Appendix as a more forceful and vivacious idea, with the terms ‘force’ and ‘vivacity’ often being supplemented by a list of others (e.g. ‘solidity’, ‘firmness’, ‘steadiness’) that are not obviously synonymous, either with one another or with ‘force’ or ‘vivacity’. The bare fact that Hume described belief in these different ways does not pose a problem as long as the different descriptions can all be integrated.11 But Hume seems to have become worried that the frequent and

109

9780826443595_Ch04_Final_txt_print.indd 109

2/9/2012 9:39:13 PM

HUME’S ACCOUNT OF PROBABLE REASONING prominent description of belief as a more forceful and vivacious idea had ‘not been so well chosen, as to guard against all mistakes in the readers’ (THN App. 1 / 623). Perhaps this was because he found readers inclined to take ‘force and vivacity’ to refer to some qualitative feature of the object that is conceived, like brightness or distinctness, rather than, as he had all along wanted to insist, a ‘manner’ in which we conceive this object, specifically, a conception with focused attention, aroused passion, and an impetus to deliberation and action – these being ‘dispositions’ of the mind that are ‘felt’ even when not acted upon and so not made evident to others. In the Appendix and insertions to the Treatise proposed in the Appendix Hume stressed that by ‘force and vivacity’ he had meant conception of an object in this ‘manner’ – conception attended with these dispositions. But Hume also confessed that he found ‘a considerable difficulty in the case; and that even when I think I understand the subject perfectly, I am at a loss for terms to express my meaning’ (THN 1.3.7.7 / 628). What may have bothered him was that appealing to ‘dispositions of the mind’ to explain belief does not sit well with what he was to go on to say about the nature of minds and mental acts (in THN 1.4.6 / 251–63 and 1.4.5.26–7 / 244–6). This pushed him in the direction of taking belief to be a feeling (presumably, the feeling of having one’s interest aroused, one’s passions elevated, and one’s inclinations determined), and in turn prompted worries about whether the feeling might be separable from the conception (denied over THN, App. 4–8 / 625–7, but affirmed at EHU 5.11 / 48). But that Hume took the presence of the dispositions to be what is ultimately constitutive of belief, and the feelings of being so disposed to be merely introspective evidence for the presence of the

dispositions – and, importantly, evidence that is defeasible – is suggested by THN 1.3.9.13 / 112–13 and 9.14 / 113–14.12 Whatever frustrations Hume may have had with his efforts, it is clear that he meant to reject the view that to believe is to perform an act of assenting to a proposition, where a proposition involves asserting a relation between two or more ideas. In particular, to believe that something exists is not to assent to a proposition joining the idea of that thing to the idea of existence. Hume rejected this possibility by arguing that we have no idea of existence distinct from whatever particular thing we conceive to exist. To conceive something as existing is no different from conceiving it (THN 1.2.6.2–6 / 66–7, 1.3.7.2 / 94–5, THN App. 2 / 623–4). He offered the ineligibility of this account of belief as a further reason to accept the alternative that to believe is just to sense or remember or be instinctively inclined to form a more vivacious idea. In a footnote, Hume went so far as to describe the division of the acts of understanding into conception, judgement and reasoning, and the definitions given of these operations as a ‘remarkable error’. These three acts of the understanding ‘all resolve themselves into the first, and are nothing but particular ways of conceiving our objects’ (THN 1.3.7.5n / 96n). This is an overstatement, since Hume did recognize that we do things like comparing objects with one another to discover relations between them, or executing arithmetical demonstrations in which one thing is inferred from another by appeal to a relation between the two. Indeed, as will be noted below, Hume went so far as to recognize a class of ‘oblique’ or ‘explicit and indirect’ causal inferences that are demonstrative in the classic sense. These operations satisfy the definition of judgement as the ‘separating or uniting of different

110

9780826443595_Ch04_Final_txt_print.indd 110

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING ideas’, and of reasoning as the ‘separating or uniting of different ideas by the interposition of others, which show the relation they bear to each other’ (THN 1.3.7.5n / 96n). The received definitions of judgement and reasoning apply to those operations that are constitutive of knowledge in the demonstrative sciences, particularly arithmetic, and knowledge of intuitive truths, such as that orange is more like red than green.13 They even apply to many of the judgements and arguments found in the empirical sciences. But they do not describe all of those operations that are constitutive of belief in the empirical sciences or in everyday life. In particular, they do not describe the most fundamental of those operations. We form fundamental beliefs neither by discerning relations between ideas nor by inferring one idea from another by appeal to an intermediate relation. Instead, we form fundamental beliefs by having lively conceptions given to us in sensation and memory, or by being instinctively compelled to form lively conceptions as a consequence of association with what is sensed or remembered. The latter is ‘not only a true species of reasoning, but the strongest of all others’ (THN 1.3.7.5n / 96n).

4. THE ARGUMENT OF THE ABSTRACT AND THE ENQUIRY Hume came to be dissatisfied with the rambling, quasi-meditative path of discovery he had dragged his reader down when presenting his ‘system’ in the Treatise. The Abstract and the Enquiry offer a far more elegant presentation of the same theory.14 They replace the opening discussion of the foundations of probable reasoning with the question of how we reason concerning ‘matters

of fact’ or existence. Like the Treatise, they leap to the conclusion that this can be done only by means of causal inference. Unlike the Treatise, they do not proceed to analyse the causal relation in terms of a problematic notion of necessary connection. The Abstract analyses causality in terms of contiguity in space, succession in time and a constant contiguity and succession in like instances, making no mention of necessary connection. The Enquiry offers no analysis of the causal relation at all, though there are passing references to a ‘supposed’ tie or connection between cause and effect (EHU 4.4 / 26–7, 4.10 / 29). In both works, necessary connection, which had played such a large role in the Treatise, comes up for discussion only after the two parts of the ‘system’ have been presented. Rather than investigate the notion of necessary connection, Hume directly proceeded to ask how causal reasoning enables us to infer the existence of unperceived objects. He first claimed that we cannot do this in advance of experience, by reference to anything we can find in those objects we consider to be causes and effects. In contrast to the scattered, sketchy, unconvincing, and ill-placed arguments of Treatise 1.3.1.1, 1.3.2.5, and 1.3.6.1 his conclusion was now justified by two different lines of argument. First, Hume appealed to everyone’s introspection, assisted by appeal to cases such as that of encountering an object for the first time or that of the biblical Adam, newly created with fully functioning, adult cognitive capacities, but no experience. Just as Adam would be unable to say what the effect of any given cause would be prior to experience, even the effect of the motion of one billiard ball towards another, so we find ourselves unable to say what the effect of a cause will be or what the cause of an effect was in novel cases (Abs. 11 / 650–1; EHU 4.6–7 / 27–8). If we think we

111

9780826443595_Ch04_Final_txt_print.indd 111

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING do perceive causal powers in more familiar cases, it is only because we have forgotten what it was like to experience these things for the first time (EHU 4.8 / 28–9). While we do often anticipate how events will turn out in novel situations (scientific experiments being the prime example), the demonstrations that we employ when doing so appeal to fundamental causal rules (cohesion, gravitation, communication of motion by impulse, etc.) that are not intuitively or demonstrably obvious, which begs the question of how we have obtained the idea of these fundamental causal relations (EHU 4.12 / 30–1). Second, Hume appealed to variations on the argument of THN 1.3.6.1 / 86–7: • According to Abs. 11 / 650–1, effects ‘follow’ from causes. Consequently, given any cause existing at one time, we can conceive any other object to exist at the following time. But when something is demonstrable, the opposite is contradictory and inconceivable. • According to EHU 4.8–11 / 28–30, every effect is a different event from its cause. Consequently, when conceiving the one, it is not necessary that we also conceive the other. If we do conceive them together, we are conscious that nothing compels us to do so, so that the conjunction is effectively arbitrary. But this means that there can be no demonstration of effects from causes or causes from effects, again because where there is a demonstration the opposite is inconceivable. The version of the argument in the Abstract omits reference to the power of the imagination to separate different objects, but it obviously rests on that assumption and it also makes more explicit appeal to the principle that ‘whatever we conceive is possible, at least in a metaphysical sense’ (Abs. 11 / 650). Like

the earlier argument at THN 1.3.6.1 / 86–7, this appears to rule out the possibility of there being any such thing as a necessary connection between the two. In contrast, the version of the argument in the Enquiry omits any reference to conceivability as a criterion of metaphysical possibility. Having established that reasoning from causes and effects is based on past experience, the Abstract proceeds to argue that any reasoning from that experience would have to depend on the principle that the future will be like the past. But (i) this principle is not demonstrably true, since a change in the course of nature is conceivable. And (ii) any attempt to prove that it is most likely true would run in a circle, since we could only appeal to the fact that it has been true in the past to argue that it will most likely continue to be true in the future. Even were we to take a constant conjunction in past experience to be evidence of the existence of a power in causes to bring about their effects, we only perceive the sensible qualities of bodies, and we can have no assurance that like sensible qualities will continue to be conjoined with like powers.15 The Enquiry mounts the same argument, but addresses it to a different question – not the question of why we suppose that the future will be like the past, but the question of why we suppose that like objects contain like hidden powers (EHU 4.7 / 28). This is not an innovation, because the same question had been raised in the Treatise (at 1.3.6.8–10 / 90–1) and the Abstract (at 15 / 652), though only as an afterthought,16 and reference to a supposition that the future will be like the past does come up in the Enquiry over the subsequent course of the argument (at 4.19 / 35–6). In addition to giving the usual reasons for a negative answer Hume also offered a new argument: since peasants, infants and animals are able to infer effects from causes,

112

9780826443595_Ch04_Final_txt_print.indd 112

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING either they do not do so by means of any argument or demonstration, or only by means of the simplest and most obvious of reasons. Yet, unless Hume was more obtuse than a peasant or child, there are no such reasons.17 Having raised these ‘sceptical doubts about the operations of the understanding’ the Abstract and the Enquiry proceed to offer a ‘sceptical solution of these doubts’ – the same, two-part solution that was presented in the Treatise. First, our experience of what has customarily been the case in the past trains habits of thought into us, so that we naturally expect the same sorts of things to happen in the future. The expectation is not rationally justified, but naturally induced. Again, the Enquiry adds a new and compelling argument for this conclusion: attributing the inference to habit offers the only plausible explanation for how it is that we come to draw a conclusion from many experiments that we would not draw after seeing just one. Secondly, we do not just infer causes and effects from one another but believe the absent partner to exist (at the contiguous place and the appropriately prior or posterior time).18 Because this belief does not arise from reasoning, but from habit, it is declared to be due to ‘a species of natural instincts, which no reasoning or process of the thought and understanding is able, either to produce, or to prevent’ (EHU 5.8 / 46–7). The Abstract and the Enquiry go on to examine what belief is, reaching the same conclusions as the Treatise and, in the Enquiry, even stating them by means of an extended quotation from the Treatise (EHU 5.12 / 48–50, quoting THN 1.3.7.7 / 628–9 with minor modifications). Interestingly, in the Enquiry these further details about belief are reserved for a distinct part of Enquiry 5, prefaced by a remark dedicating the part to ‘such as love the abstract sciences, and can be

entertained with speculations, which, however accurate, may still retain a degree of doubt and uncertainty (EHU 5.9 / 47)’. ‘Readers of a different taste’ are told that the part may be neglected without impairing an understanding of subsequent portions of the book. Part 2 of Enquiry 5 (EHU 5.10–22 / 47–55) is nonetheless important. As Hume stressed in the same breath in which he advised ‘readers of a different taste’ to move on, delving into the question of what belief is and how it arises will uncover ‘explications and analogies that will give satisfaction’. The ‘satisfaction’ Hume had in mind is not just the satisfaction of idle intellectual curiosity, but the satisfaction of objections to the account of belief laid out in the concluding paragraphs of the Enquiry 5, Part 1. A concern with uncovering ‘analogies’ between the account of belief and other operations of the mind is a constant of Hume’s thought about his ‘system’ (cf. THN 1.3.8.3 / 99; App. 9 / 627; Abs. 23 / 655), for good reasons that have already been alluded to. There is very little positive argument to justify Hume’s ‘system’. The meditative path of discovery of the Treatise offers only rhetorical support,19 and the Abstract’s and Enquiry’s effective reconstruction of that path into a critique of the view that our causal inferences are justified by appeal to facts and rules at best put Hume in a position to claim that causal inference is not based on reasoning, not to claim that it is based on a habit of association and a transfer of belief from an impression or memory. The same can be said of a new argument, presented only later, to the effect that reasoning is too slow and uncertain in its operations to be entrusted with an operation as important for survival as causal inference (EHU 5.22 / 55) and of the Enquiry’s appeal to the abilities of peasants, children and animals to draw causal inferences. The one

113

9780826443595_Ch04_Final_txt_print.indd 113

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING positive argument for the theory presented so far is the Enquiry’s appeal to the problem of how we draw a conclusion from repeated experiments that we cannot draw from just one, and that argument offers a justification of the least satisfying sort: inference to the best explanation. Hume hoped that by uncovering ‘analogies’ between belief and other operations of the mind he would be able to offer a more compelling, Newtonian argument by induction from the phenomena to a general rule. The general rule would provide justification for the two-part system, as a special case, but it would in turn be supported by induction from all the analogous cases revealed by experience. However, Hume had come to think that the public had no taste for this sort of investigation, particularly if drawn out to any great length (Abs. Preface, 1–2 / 643). His solution was to drastically abbreviate the argument, focusing just on the exposition of analogous cases, and to invite impatient readers to skip ahead. In the Treatise he went on at much greater length, not only identifying analogous cases, but worrying about contrary evidence, refining the system to account for it, and appealing to the system to account for a wide range of other phenomena, thus adding a demonstration of explanatory power to the other reasons for accepting the system. Because the Enquiry merely repeats some of what the Treatise had to say on this score,20 I focus on what Hume had to say in the Treatise in what follows.

5. ANALOGIES, EXPERIMENTS, RECALCITRANT DATA, AND REFINEMENTS In both the Treatise (1.3.8.3–5 / 99–100) and the Enquiry (5.15–18 / 51–3) – the relevant

passages are identical in both) – Hume noted the following ‘analogies’ between causal inference and other operations of the mind: the picture of an absent friend ‘enlivens’ the idea of that friend, as well as the passions that idea occasions; the ‘mummeries’ of the Roman Catholic religion ‘enliven’ devotion; ‘sensible types and images’, which have a greater influence on the fancy than any other, ‘convey that influence’ to the ideas they resemble; objects that are placed in the vicinity of other objects ‘transport’ the mind ‘with a superior vivacity’ to ideas of those other objects (e.g. passing the house next door on my way home gives me an idea of my home that ‘imitates an immediate impression’). Importantly, in all of these cases the trigger (the picture, the ceremony, the icon, the neighbouring objects) must be both experienced and ‘naturally related’ to the target (in the cases mentioned, by relations of resemblance or contiguity); if the trigger is merely imagined, the target is not enlivened; if the trigger is unrelated, the idea of the target does not even arise. The case is the same with causal inference, as Hume proceeded to prove by appeal to three experiments (THN 1.3.8.8–11 / 101–3): suppose the natural relation (in this case of constant conjunction) is absent (as, for example, when experiencing a cause for the first time). Then the associated idea does not arise. Now suppose a constant conjunction between the trigger and the target has been experienced in the past. Then, solely for that reason and without the assistance of any intermediate process of argument or justification or appeal to general rules, a vivacious idea of the target arises upon experience of the trigger. Now suppose the trigger is not experienced but only imagined. Then the associated idea does arise, but it has no vivacity. By induction from all these phenomena, Hume declared it to be ‘a general maxim

114

9780826443595_Ch04_Final_txt_print.indd 114

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING in the science of human nature, that when any impression becomes present to us, it not only transports the mind to such ideas as are [naturally]21 related to it, but likewise communicates to them a share of its force and vivacity’ (THN 1.3.8.2 / 98). This maxim was no sooner justified than Hume acknowledged a difficulty. He had defined belief to be nothing but a more ‘vivacious’ idea. But he had also maintained that belief only arises from causal inference, not from the other natural relations of resemblance and contiguity, notwithstanding that, according to the maxim, they all enliven ideas. The three claims are inconsistent (THN 1.3.9.2 / 107). At this point Hume stood on the brink of momentous discoveries, presented with an opportunity to reassess his earlier, ill-considered position that belief in matters of probability can only arise from causal inference. He had discovered that basic causal inferences are inferences from the constancy of temporal succession in resembling cases. But we also draw inferences from the constancy of spatial arrangement in resembling cases. Quite apart from forming any background beliefs about the causes of the immobility of landmarks, we rely on the constancy of the position of houses, trees, the pole star and other geographical or astronomical objects to navigate, and when we do so we reason from experienced objects to their unperceived surroundings, not to their unperceived causes or effects.22 Hume himself recognized this without realizing it when he wrote: Suppose I see the legs and thighs of a person in motion, while some interpos’d object conceals the rest of his body. Here ’tis certain, the imagination spreads out the whole figure. I give him a head and shoulders, and breast and neck. These members I conceive and believe him to be possessed of. (THN App. 4 / 626)

Since the head, shoulders, breast and neck are neither causes nor effects of the legs and thighs, the inference here is not causal, even though Hume recognized that it produces belief. The same holds of Hume’s description of ‘our approach to any object; tho’ it does not discover itself to our senses’ as leading that object to operate on the mind ‘with an influence that imitates an immediate impression’ (THN 1.3.8.5 / 100). An ‘influence that imitates an immediate impression’ is just a belief. The case of resemblance poses more of a problem. Seeing the son of a long-dead friend (EHU 5.19 / 53) does not produce a belief in the existence of an unperceived object. At best, it rouses old memories and enlivens the associated passions.23 One reason for this is that the resemblance relation calls up an idea of a resembling object without giving any further indication of where that object is placed in space and time, leaving us with no inclination to ascribe it a location in the real world. Our causal and geographical inferences, on the other hand, involve not just association of objects, but association with places and times where the object is to be found. Not surprisingly, therefore, when resemblance is bound up with relations of time and place, it has the same influence as constant succession in time and constant conjunction in space. This is most notably the case with our beliefs about the identity of objects over time, where we suppose a continuum of intermediate states to exist unperceived between observed, resembling, earlier and later states. This is just a sketch of how Hume might have gone on to investigate the possibility that probabilistic reasoning involving all three of the ‘inconstant’ relations of causality, contiguity and identity could be grounded in the three ‘natural’ relations of constant succession in time, constant conjunction in space and closest resemblance at contiguous places over time.

115

9780826443595_Ch04_Final_txt_print.indd 115

2/9/2012 9:39:14 PM

HUME’S ACCOUNT OF PROBABLE REASONING Unfortunately, Hume did not take this path. (THN 1.4.2.15–23 / 194–9 is perhaps the most lamentable consequence of that decision.) He did go so far as to declare that we take our sense experiences and memories to constitute a ‘system’ of ‘realities’, and that we join a second ‘system’ to it, consisting of the unperceived causes or effects of these ‘realities’ (THN 1.3.9.3–4 / 108). But he never paused to consider that what makes the objects of the senses and memory a ‘system’ is that each is related to all the others in virtue of its unique location within a single space and time, and that what enables the causal relation to augment the system is that it directs us where to localize unperceived objects in this space and time – something that constant contiguity and resemblance insofar as it is bound up with identity relations could also do. Instead, Hume maintained that because any given object resembles and is spatially contiguous to a huge variety of other objects, the mind senses a certain ‘caprice’ or feeling of liberty in making the association with just one. This feeling of liberty prevents the easy and unnoticed transition from one object to another that Hume had earlier identified as essential for the transfer of the mental dispositions characteristic of belief. It also introduces new feelings of ‘looseness’ and ‘weakness’ that are contrary to the feelings of stability and strength characteristic of belief. Moreover, any tendency we might have to include objects thought of under these conditions in the system of ‘realities’ would produce repeated experiences of having our expectations disappointed. As a consequence, we would learn to associate objects thought of under these conditions with fictions (THN 1.3.9.6 / 109–10).24 Causal relations are very different. Any given object is related to just one other object as its cause and just one other object as its effect, so there is no ‘looseness’ to the association.25

These reflections mandate a revision to Hume’s maxim, although he never said so. While impressions may transport the mind to any ideas that are naturally related to them, they only communicate a share of their force and vivacity sufficient to induce belief to ideas of those objects that have been customarily conjoined with them in the past. This revised maxim explains an attempt at ‘confirmation’ of the ‘hypothesis’ that Hume made at THN 1.3.9.16–19 / 115–17. The attempt appeals to an example that would otherwise serve more to falsify than to confirm the hypothesis. If the repetition of a conjunction (making it ‘customary’) plays a more important role in producing belief than the force and vivacity of the impression, and if belief is a mental ‘disposition’ involving things that can be produced merely by repetition, such as fixed attention, familiarity and stability of the object, then we should expect that belief could arise from the mere repetition of an idea even in the absence of association with an impression or memory. Hume considered this in fact to be the case, most notably with the beliefs produced by education, which he considered to provide outstanding confirmation for the hypothesis because, as he claimed, education is responsible for more than half our opinions and is more influential than either abstract reasoning or experience (THN 1.3.9.19 / 117). He also instanced the tendency of amputees and the bereaved to be unable to accept their loss, and of people to consider themselves to be on intimate terms with personages they have only read about. It is hard not to wonder about the aptness of these examples or the soundness of Hume’s implicit view of how education produces belief, but people do have a tendency to believe what they hear from those around them simply because everyone is saying it and even though no one

116

9780826443595_Ch04_Final_txt_print.indd 116

2/9/2012 9:39:15 PM

HUME’S ACCOUNT OF PROBABLE REASONING is in a position to testify to the truth of what they are saying (the belief in an afterlife being an outstanding example). Besides this appeal to a confirming experiment, Hume justified the hypothesis by appeal to its explanatory power (THN 1.3.9.9–15 / 110–15). The hypothesis is able to explain such things as (i) why pilgrimages strengthen belief; (ii) why it is wrongly supposed that the communication of motion by collision could be anticipated in advance; (iii) why we have a much more vivid conception of the vastness of the ocean from vision than from hearing; (iv) why we are credulous, even in the face of contrary experience; (v) why we cannot take the infinite rewards and punishments of an afterlife seriously, even if we believe in them; and (vi) why we enjoy religious discourses and dramatic performances that excite the disagreeable passions of fear and terror. In all of these cases the explanation is the same. Though the natural relations of contiguity and resemblance are not able to produce belief on their own, when a belief has once been formed, so that there is no sense of ‘caprice’ in its conception, it will be further enhanced by relations of contiguity (i) and resemblance (ii–iv) holding between the impression and the idea, but also weakened by the opposite relation of dissimilarity (v), with the weakening of belief in turn accounting for (vi).

6. THE PSYCHOLOGICAL FOUNDATIONS OF EPISTEMOLOGICAL NORMATIVITY Taken together, the two parts of Hume’s ‘system’ would appear to rule out any role for logic in probabilistic inference – any role for the conscious, deliberate application of

demonstrably truth-preserving or probability-preserving rules to draw inferences from the observed to the unobserved. We do not discern a relation between causes and effects by comparing them with one another and then appeal to this relation to draw inferences to the unobserved. Instead, we are instinctively impelled to form ideas of objects of a sort that have in the past been frequently observed to be constantly conjoined with currently sensed or remembered objects, doing so in the absence of any memory of those past occasions or conscious inference from them (THN 1.3.8.13 / 103–4). And we do not judge that these objects must exist but are instinctively impelled to form a more ‘lively’ conception of them – a conception that ‘gives them more . . . influence; makes them appear of greater importance; infixes them in the mind; and renders them the governing principles of all our actions’ (THN 1.3.7.7 / 629). These are results that Hume trumpeted in both the Treatise (1.3.8.12 / 103) and the Enquiry (5.8 / 46–7), writing that because objects have no discoverable connection with one another, we can only draw an inference from the one to the other with the aid of ‘custom operating on the imagination’, and that ‘all probable reasoning is nothing but a species of sensation’ (THN 1.3.8.12 / 103), so that when we prefer one argument to another we do nothing but decide from our feelings concerning the superiority of their influence, meaning that belief is ‘a species of natural instincts, which no reasoning or process of the thought and understanding is able, either to produce, or to prevent’ (EHU 5.8 / 46–7). In the Enquiry, Hume pretended that even though our beliefs are not drawn from observation in accord with truth-preserving or probability-preserving rules, there is a ‘preestablished harmony between the course of

117

9780826443595_Ch04_Final_txt_print.indd 117

2/9/2012 9:39:15 PM

HUME’S ACCOUNT OF PROBABLE REASONING nature and the succession of our ideas’ ensuring that beliefs will be produced in us in tandem with the way causes and effects succeed upon one another in nature, and providing ‘those who delight in the discovery and contemplation of final causes’ with ‘ample subject to employ their wonder and admiration’ (EHU 5.21 / 54–5). This is a singular instance of misdirection, inconsistent with the candour that is otherwise characteristic of his work. It is not just that, on his account, there could at best be a pre-established harmony between the past course of nature so far as it has been observed by us and the succession of our ideas. Hume’s account entails that there should not even be that much. According to Hume’s theory, belief is a more vivacious idea resulting from association with an impression or memory. It therefore depends on the original vivacity of the impression or memory and the strength of the associative link between the impression or memory and the idea. If either of these is weakened, the belief will be as well. But, as Hume observed in the Treatise, an impression is more vivacious than a memory, and a recent memory more vivacious than an older one. A recent observation of a conjunction between types of events also produces a stronger disposition to associate those events than an earlier one. As the Hume of the Treatise went on to admit, these factors entail that the course of our ideas should not be in harmony with the past course of nature. Instead, it should be disproportionately influenced by the most recent observations (THN 1.3.13.1–2 / 143–4). This is not all. Hume further observed that the strength of association is also affected by the ease with which it is made, so that a causal inference that needs to be drawn by appeal to a number of intermediate causes should be less strongly believed

than one that is more immediate (THN 1.3.13.3 / 144). And just as there are factors that lead us to overlook or discount connections found in the past course of nature, so there are factors that lead us to suppose the existence of connections that were not there. According to the theory, we are disposed to consider similar objects to have similar causes or effects. But any given object is a compound of many different characteristics, only one of which may be constantly conjoined with a cause or effect. Even if we have learned to distinguish the essential characteristic from the superfluous ones, when we encounter an object that resembles a cause or effect only in superfluous ways, the resemblance should lead us to conceive the associated effect or cause, and the ease of the association together with the vivacity of the encountered object should induce a kind of bigoted or prejudicial belief, which holds sway despite our recognition of the superfluity of the resemblance (THN 1.3.13.7 / 146–7 and 1.3.13.9 / 147–8). Nor are these the only such cases. Hume noted that education ‘not only approaches in its influence, but even on many occasions prevails over that [belief] which arises from the constant and inseparable union of causes and effects’ (THN 1.3.9.17 / 116). He further noted that we do not regulate ourselves entirely by experience of the governing principles of human nature when deciding whether to believe testimony, but instead ‘have a remarkable propensity to believe whatever is reported, even concerning apparitions, enchantments, and prodigies, however contrary to daily experience and observation’ (THN 1.3.9.12 / 113). This is in part due to the influence of the resemblance relation between ideas (supposed to exist in the minds of others on the basis of their words) and facts, which strengthens the associative

118

9780826443595_Ch04_Final_txt_print.indd 118

2/9/2012 9:39:15 PM

HUME’S ACCOUNT OF PROBABLE REASONING relation beyond what is warranted by experience of their constant conjunction. But it is also due to the fact that ideas that arouse passions are reciprocally enlivened by those passions (THN 1.3.10.4 / 120). This looseness of fit between the course of our ideas and the past course of nature is not necessarily a bad feature of Hume’s ‘system’. As a matter of fact, people’s beliefs are more strongly influenced by recent experiences; people are less inclined to lend credence to the conclusions of complex arguments; and people are disposed to bigotry, blind adherence to received opinions, and credulity. The fact that the course of our ideas fails to track the past course of nature in just these ways is further confirmation that Hume’s theory has correctly captured the psychological mechanisms responsible for human belief formation and is a further instance of its explanatory power in accounting for those inferences we are in fact psychologically compelled to draw. But this is still not an entirely happy result. While many of us form blinkered, prejudicial, obstinate and credulous beliefs, we do not all do so. At least, we do not all do so all of the time. At the very least, we do not all think we should do so, even if, as a matter of fact, we find ourselves irresistibly compelled to do so anyway. As Hume himself remarked, inferences skewed by recent experience are not ‘receiv’d by philosophy as solid and legitimate’ because ‘philosophers’ do not think that the same event or the same conjunction between events should provide less evidence a month from now than it provides today (THN 1.3.13.1 / 143). Furthermore, education is not ‘recogniz’d by philosophers’ because it ‘is an artificial and not a natural cause’, and because ‘its maxims are frequently contrary to reason, and even to themselves in different times and places’ (THN 1.3.9.19 / 117). And credulity is, as Hume himself admits,

a universal and conspicuous ‘weakness of human nature’ (THN 1.3.9.12 / 112). This raises two problems. If all beliefs are ultimately unjustifiable and all are founded on the same operation of a transmission of vivacity across associative links, how could any of us have come to think that some of them are better than others? And how could some of us (e.g. a gambler who places bets in accord with a calculation of the probability of outcomes, ignoring the results of recent games) not only think that certain beliefs are better than others but manage to form their own beliefs accordingly? Hume had solutions for both problems.

7. EPISTEMOLOGICAL NORMS In all cases we transfer our experience to instances, of which we have no experience, either expressly or tacitly, either directly or indirectly. (THN 1.3.8.14 / 105) According to the account that has so far been presented of Hume’s ‘system’, causal inference is an unconscious (‘tacit’), instinctive (‘direct’) operation resulting from, not carried out in cognizance of, past experience. On the first few occasions of observing objects of one sort to be followed by objects of another sort, we are unimpressed, and not disposed to draw any inference when encountering objects of either sort in the future. But as we make more and more observations of the conjunction of the two objects, we develop a habit to think of the one when presented with the other. As the number of observations increases, the habit strengthens, and as the habit strengthens, more and more of the mental dispositions characteristic of belief and attendant upon experience and memory

119

9780826443595_Ch04_Final_txt_print.indd 119

2/9/2012 9:39:15 PM

HUME’S ACCOUNT OF PROBABLE REASONING come to attend the associated idea. Belief, therefore, is something that comes in degrees, varying from conjecture to certainty in proportion to the strength of the habit and hence to the number of past observations up to the point where a sufficient number of observations have been made to produce a habit that mimics experience in its effects (THN 1.3.12. 2 / 130–1). Importantly, we do not recall the past observations or appeal to them to justify our belief. The past observations have made us develop a habit and that habit alone produces the belief (THN 1.3.8.13 / 103–4). But there are twists to this simple story. One twist arises from the fact that past experience is not always uniform.26 Sometimes, objects of one sort are not always followed by objects of another sort. When that happens, the contrary experiences weaken the habit. Over time, we end up with a habit that would be strong or weak in proportion to the number of confirming instances in the total number of trials, but for the influence of the ‘unphilosophical’ factors mentioned above (THN 1.3.12.6 / 132–3). The belief we get from inconstant experience is still ‘tacit’ and ‘direct’. But in Treatise 1.3.12.2 / 130–1, Hume declared that there are other kinds of causal inference. The kind based on the gradual development of a habit over the course of a uniform past experience is not, he claimed, to be found in anyone ‘who is arriv’d at the age of maturity’ (THN 1.3.12.3 / 131), and the kind based on inconstant experience is one that ‘we have but few instances of in our probable reasonings’ (THN 1.3.12.7 / 133). Mature adults are supposed to draw causal inferences after just one experience, supposing it has been obtained in circumstances where ‘all foreign and superfluous circumstances’ have been removed. And Hume maintained that when the conjunction between causes and effects

is not entirely uniform, we seldom rely on a gut reaction but instead deliberately recall the past experiments, count up the number of confirming and contrary instances, and form beliefs that are stronger or weaker in accord with a mathematical calculation of probabilities.27 These are extraordinary claims that at first sight seem incompatible with the ‘system’. A habit cannot be formed after just one experience, and a mathematical result is based on the perception of a relation between ideas that have no vivacity, and so should not produce belief.28 Hume’s account of how we manage to do these things lays the foundations for epistemological norms and an account of action in accord with those norms.

8. GENERAL RULES Hume claimed that while we do not as a matter of fact recall any past experiences when drawing inferences concerning conjunctions of causes and effects that have been constantly observed since infancy (e.g. stones fall, fire burns, water suffocates), we do ‘assist the custom and transition of ideas’ by recalling past experiences when encountering more rare or unusual objects (THN 1.3.8.14 / 104). There is nothing about the system that would suggest we are prevented from doing so. On the contrary, similar objects can jog the memory as well as the imagination, particularly in unusual cases. And we can be motivated to recall past instances by passions such as curiosity, love of fame, fear and hope. Our causal inferences could, therefore, be sometimes ‘express’ rather than ‘tacit’. Just as nothing prevents us from expressly recalling and reviewing past experiences, so

120

9780826443595_Ch04_Final_txt_print.indd 120

2/9/2012 9:39:15 PM

HUME’S ACCOUNT OF PROBABLE REASONING nothing prevents us from drawing conclusions from those past experiences ‘indirectly’, by explicit appeal to causal rules, learned from past experience. The chief such rule is the general one that like objects, placed in like circumstances, will have like effects (THN 1.3.15.6 / 173). This rule is not justified by past experience – no causal rule is. But past experience does lead us to form it and believe it. It is ‘merely habitual’ as Hume put it (THN 1.3.8.14 / 104–5). Once formed, it can be expressly appealed to in order to justify causal inference from as little as a single past experience (THN 1.3.8.14 / 105). This accounts for why people who have reached the age of maturity will draw causal inferences after just one experience rather than needing to be trained by a number of experiences. ‘Express and indirect’ causal inferences can be based on other rules besides the rule that similar objects placed in similar circumstances will have similar effects. As has already been noted, any striking resemblance between an unfamiliar object and a familiar one will lead us to suppose that the unfamiliar object has the same causes or effects as the familiar one, even if there are some dissimilarities between the objects and the attendant circumstances are not quite the same. This is the foundation of prejudice. As we grow older, we encounter cases in which our prejudices are disconfirmed and reflection on these cases leads us to appreciate the importance of distinguishing between superficial characteristics, which are often but not always present in causal conjunctions, and essential ones, which are always present (THN 1.3.13.11–12 / 149–50). We begin to think that it is not good enough to suppose that any object that bears any striking resemblance to objects we have experienced before will have the same causes or

effects. Instead, we need to be sure that we have eliminated ‘all foreign and superfluous circumstances’ as Hume put it (THN 1.3.8.14 / 104). We first consider it to be at least possible that where two resembling objects have different effects, all the similarities between them must be foreign and superfluous and the different effects must arise from some hidden respect in which the causes differ. As it turns out, our experience of regularly finding these hidden features upon a more exact scrutiny habituates us to that belief (THN 1.3.12.5 / 132). A similar course of experience habituates us to the belief that where strikingly different objects have the same effect, all the differences between them are foreign and superfluous and the effect must arise from some circumstance common to the two. We are further habituated to believe that any increase or diminution in the effect must be due to a compound cause, and that any cause that persists over time before being followed by its effect cannot be the sole cause of that effect (THN 1.3.13.11–12 / 149–50, 1.3.15.7–10 / 174–5). These ‘rules by which to judge of causes and effects’ justify other causal inferences. Taken together, they constitute a logic of causal inference (THN 1.3.15.11 / 175) – a system of rules that can be expressly appealed to in drawing indirect or ‘oblique’ causal inferences. This is one part of the answer to the question of how Hume could provide for a logic of causal inference.29 Once we have come to form and accept the rules, inferences that are in accord with them will be approved of as wise or justified, whereas those that violate them will be condemned as foolish. This will be the case even if the person making the normative assessment is personally unable to follow the rules due to the influence of ‘unphilosophical’ factors.

121

9780826443595_Ch04_Final_txt_print.indd 121

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING 9. PROBABILITIES Hume noted that while past experience may have habituated us to the thought that same objects have same effects, we do not always find this to be the case. Contrary experience does not, however, lead us to diminish our confidence in the general rule because we have been further habituated to believe that where same objects have different effects they will be found upon more exact scrutiny to differ in some previously unnoticed way. We may not always have the opportunity, the curiosity, the leisure, or the resources to search for this hidden circumstance. But this just means that, in the absence of a perception of a previously hidden cause, our habits will lead us to believe that it must exist. The case here is similar to what Hume ought to have said concerning the spatial contiguity of causes and effects. If we fail to observe a succession of cause and effect because closing our eyes or turning our heads leads us to fail to observe the contiguous regions, then we do not think we have observed a failure of the expected succession to occur. On the contrary, this is exactly the sort of case in which the habit kicks in and leads us to form a belief in the unperceived existence of the cause or effect at the contiguous location. The only difference between the two cases is that this time Hume did not fumble, as he did at Treatise 1.4.2.21 / 197–8, but declared that, ‘from the observation of several parallel instances, philosophers form a maxim, that the connexion betwixt all causes and effects is equally necessary, and that its seeming uncertainty in some instances proceeds from the secret opposition of contrary causes’ (THN 1.3.12.5 / 132). But, Hume went on to observe, even though we may believe there is a hidden cause, in the absence of opportunities, curiosity, leisure

or resources to search for it, we have no recourse but to consider those circumstances that tend to accompany it to be signs of its likely existence, and so to take apparently similar objects to have similar effects, albeit with a diminished degree of certainty proportioned to our experience of the degree of regularity in the connection between occurrences of the superfluous circumstances and the effect (THN 1.3.12.6 / 132–3). These inferences are ‘probable’ in the strict sense of being something less than ‘entirely free of doubt and uncertainty’ (THN 1.3.11.2 / 124). Importantly, they are not ‘tacit and direct’. They do not result from a habit that has been strengthened or weakened by past experience – but also by a host of ‘unphilosophical’ factors. They are instead ‘oblique’ or explicit and indirect. They result from a survey of past instances and a calculation of the proportion of confirming experiments in the total number of trials. It has already been noted why we might be impelled to remember past experiences and want to survey them. Hume’s remaining challenge in accounting for our ‘explicit and indirect’ reasoning concerning matters of strict probability was not to come up with a mathematical theory of the calculation of probability. It was to explain how we come to proportion belief in accord with any merely mathematical calculation. The basic facts that need to be accounted for are obvious: in cases where there has never been an exception to a succession, we should form a belief that is ‘entirely free of doubt and uncertainty’; in cases where it is no more likely that an event will occur than not, we should have no belief, and in the intermediate cases we should have a proportionally strong or weak belief. So, where there is what we might call a ‘fifty-fifty chance’, we should have no belief; where

122

9780826443595_Ch04_Final_txt_print.indd 122

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING there is a ‘one hundred per cent chance’ we should have certain belief; and where there is a ‘seventy-five percent chance’ we should have a belief that is half way between indifference and certainty. In the last case, importantly, we do not form a certain belief in the proposition that the event has a probability of fifty (or seventy-five) percent;30 we form a less vivacious conception of the event – one that is half-way between a certain belief and an indifferently entertained idea as measured by the feeling that attends it and the strength of the characteristic mental dispositions. We should get a belief of this strength regardless of what our views are on how to calculate probabilities mathematically, or whether we employ any mathematics at all beyond a bare survey of instances. In explaining how the belief arises Hume distinguished between two cases, that of belief in ‘chances’, and that of belief in ‘inconstant causes’. Chance arises when a cause has an effect that can indifferently take one or another of a finite number of alternative forms, e.g. the toss of a die causes it to fall with one or other of its six sides facing up. Causes are ‘inconstant’ when either the indifference condition or the finitude condition is not met. In the first case we have been strongly habituated to associate the generic effect with the cause. So, for example, upon witnessing a die being tossed we form a strong belief that it will fall with one side facing up. But now suppose we ask ourselves exactly which side will face up. Because the generic effect has a number of equally possible, mutually incompatible forms the strong belief in the generic effect becomes divided, with an equal portion going to each alternative. But, because the alternatives are all mutually incompatible, the divided beliefs cancel one another out. We are left with a very strong belief that

the generic effect will occur, but indifferent over which form it will take. For example, the belief that the die will fall with ‘side one’ facing up would be one-sixth the strength of the belief that it will fall with some side or other facing up – but for the fact that we cannot have any degree of belief that ‘side one’ will fall face up without disbelieving that any of the other sides will fall face up. Since, however, all those possibilities are considered to be equally likely, they cancel one another out, leaving us with certainty that some side or other will face up, but indifference about which one. Hume’s account of how we form beliefs about the effects of inconstant causes grows out of a refinement of the account of belief in chances. If some of the chance alternatives resemble one another, e.g. four of the six sides have the same figure on them, a survey of the alternatives causes their portions of the original belief to combine. In that case the alternatives do not perfectly cancel one another out, e.g. the possibility that the die will land with the more familiar figure facing up receives four of the six portions, only two of which suffice to cancel the rival possibilities, leaving us with a residual belief that the more familiar figure will face up. This residual belief is two-sixths of the way between indifference and conviction.31 Belief in the outcome of inconstant causes is like this, except that in this case we do not start off with a strong belief in the occurrence of a generic effect. Hume instead supposed that we develop the habit of expecting the future to be like the past. The strong belief in this uniformity principle plays the same role as the strong belief in the occurrence of a generic effect. If in the past one egg in every crate has been rotten, we will transfer that past experience to the future in the sense that any subsequent impression or memory of a

123

9780826443595_Ch04_Final_txt_print.indd 123

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING crate of eggs will produce a strong belief in the existence of eleven sound eggs and one rotten one. On picking out any one egg from the crate and forming an idea of what we will smell upon cracking it open, that strong belief is divided into twelve packets, eleven of which resemble one another in being attached to ideas of sound eggs, and one of which is attached to the idea of a rotten egg. Since the possibilities are incompatible, we cannot believe them both to result from cracking the egg. Since one of them comes up so much more often in a mental survey of alternatives, we cannot be perfectly indifferent, either. The odd possibility is cancelled out by one of the opposing packets leaving us with a belief that this egg will prove to be sound that is ten-twelfths as strong as the original belief that there will be eleven sound eggs in the crate. There are oddities about Hume’s presentation of this account in both the Treatise and the Enquiry. The account offered by the Treatise is repeated four times over (first at THN 1.3.12.8–12 / 133–5, a second time at 1.3.12.13–18 / 135–7, again at 1.3.12.19 / 137–8, and a fourth time at 1.3.12.20–2 / 138–40). Whether this is because Hume was unsure of himself or particularly proud of his result is unclear. In Enquiry 6 / 56–9 he retreated from the attempt to provide a calculus of the strength of belief, at first attributing probabilistic belief to ‘an inexplicable contrivance of nature’ but then saying that his account of belief as a firmer and stronger conception of an object allows us to explain matters a bit further by saying that where ‘a great number of views . . . concur in one event’ they ‘fortify’ the conception of that event in the imagination and so produce belief. This reticence notwithstanding, the full theory of belief as proportioned in accord with the subtraction of less frequent

possibilities from more frequent ones reappears at EHU 10.4 / 110–11. Whatever ambivalence Hume might or might not have had about his account, he suggested in the Treatise (1.3.11.1 / 7–8) and stated outright in the Abstract (4 / 646–7) that he had offered the only adequate explanation of probable belief that had ever been given. There could be no demonstrative account of probable belief because it is in principle impossible to demonstrate that the event that is most often observed will occur (in that case, it would not be merely probable). There can be no probable account of probable belief because those who claim that a survey of past results can at least make us certain about which event is most likely to occur are actually doing no more than uttering the trivial claim that the event that has come up most often in a survey of past results is the event that has come up most often in a past survey of results, not giving us a reason to believe, with any degree of conviction, that this event will occur on any future occasion.32 This is further proof, in Hume’s eyes, that belief must be a more vivacious manner of conception rather than the product of a judgement.

10. ‘PHILOSOPHICAL’ BELIEF By all that has been said the reader will easily perceive, that the philosophy contained in this book is very sceptical, and tends to give us a notion of the imperfections and narrow limits of human understanding. (Abs. 27 / 657) Hume had no sooner presented his account of how we arrive at weaker or stronger beliefs about matters of probability than he remarked that recent experiments weigh more

124

9780826443595_Ch04_Final_txt_print.indd 124

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING than earlier ones in our assessments of the probability of outcomes, whether we realize it or not. The freshness of an experiment ‘has a considerable influence on the understanding, and secretly changes the authority of the same argument, according to the different times, in which it is propos’d to us’ (THN 1.3.13.1 / 143). He also said that this is an effect that ‘has not had the good fortune’ to be ‘receiv’d by philosophers, and allow’d to be [a] reasonable foundation of belief and opinion’ (THN 1.3.13.1 / 143). As already noted, this is just one of many ‘unphilosophical’ influences on belief, influences that we are as a matter of fact susceptible to, for reasons that Hume’s ‘system’ succeeds admirably well at explaining. We might conjecture that ‘philosophers’ are led to condemn these influences as a consequence of having been habituated to accept certain rules, such as the rule that explicit and indirect probable inferences are more often correct than tacit and indirect ones. But not everyone is habituated to accept the same rules. Hume noted that ‘the vulgar’ are not habituated to accept that same causes must have same effects or that when a cause fails of its usual effect a more exact scrutiny will uncover some previously hidden circumstance that is the true cause. Instead, their experiences have habituated them to accept that causes are not perfectly regular in their operations, even though nothing impedes them (THN 1.3.12.5 / 132). Someone with that outlook will be less inclined to recognize a distinction between superfluous and essential components of causes (hence more inclined to prejudice), and less inclined to accept that the course of nature cannot change (hence more inclined to form their beliefs on the basis of recent evidence). And a philosopher is in no position to convince them of the error of their ways since the

philosopher’s commitments have no other foundation than (at best, and not always) the past course of their own experience – a foundation that the vulgar can appeal to as well. The difference between the vulgar and the philosophers is that they have had different experiences. Each judges in accord with what their own experiences have made them and neither is in a position to appeal to their experiences as a higher authority. Nor is this the only impediment to ‘philosophical’ belief. Just because ‘philosophers’ approve of the ‘rules by which to judge of causes and effects’, it does not follow that they will always form their beliefs in accord with those rules. Very few of us have been habituated to think (1) that whenever a supposed cause fails of its usual effect a hidden cause will be discovered upon more exact scrutiny, (2) that the bare passage of time does not destroy the authority of an experiment, or (3) that beliefs that are based on recent, lurid anecdotes will more often prove to be false than those that are based on an impartial survey of cases. The reason most of us accept these and other such rules is not personal experience in the laboratory, but uncritical acceptance of the testimony of others, education or the influence of a few, recent, notable errors. That is perhaps not a bad thing. Ironically, if anything enables ‘philosophers’ to draw their inferences just in accord with the rules by which to judge of causes and effects, it is not having been habituated to accept them but having been educated to accept them.33 Education is among the most illegitimate but also the most powerful of the factors influencing our belief. Personal past experience conveys a degree of vivacity to associated ideas that is vulnerable to being artificially enhanced or diminished by resemblance, contiguity, passion or the passage of time, whereas education is largely impervious to those influences.

125

9780826443595_Ch04_Final_txt_print.indd 125

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING Hume’s position is not entirely sceptical. Although he never said so in quite so many words (but see THN 1.4.7.13 / 271–2), he would probably have agreed that a survey would show that ‘philosophical’ beliefs have more often turned out to be correct than ‘vulgar’ ones. The rules followed by ‘philosophers’ in arriving at their beliefs, therefore, have a title to be considered the logic of probable reasoning rather than just an anthropological description of the epistemological aspirations of a certain social class.34 But the rules have this title only for those already habituated to accept that the future will be like the past, and to accept that explicit and indirect probable inferences, based on an impartial survey of past cases, are to be preferred to tacit and indirect ones. Those not habituated (or educated) to accept these rules could justly reject any appeal to a survey of past instances as question-begging. And even those who do accept the presuppositions of the argument will not always be able to resist the other factors inducing them to form beliefs. ‘Unphilosophical’ belief will persist, even among those who know better. This explains the characteristic pessimism displayed in Hume’s Essays, Natural History of Religion, and History of England.35 Under certain social conditions the arts and sciences will flourish and philosophical learning will triumph over vulgar superstition. But once entrenched in a society, there is no guarantee that the arts and sciences will progress. A change in circumstances, beyond anyone’s ability to control or even predict, altering the course of people’s experiences, can change their inferential practices and the intellectual culture that had developed can be supplanted by the crudest barbarism. Hume did think that there is one means by which the influence of ‘unphilosophical’ factors on belief might be mitigated:

an experience of the force of sceptical arguments.36 He thought that someone who has once been convinced of the weakness and fallibility of our powers of knowledge will be permanently changed by that experience. Forever afterwards, they will be doubtful about all their beliefs and hesitant about forming them. This doubt and hesitancy will naturally dispose them to distrust testimony and education and refrain from peremptory (tacit, direct) judgement (EHU 5.1 / 40–1, 12.24–6 / 161–3; DNR 1.133–4). It will also extend to philosophical beliefs, as Hume made clear in the last words of Treatise 1.4.7 / 274. . . . we are apt not only to forget our scepticism, but even our modesty too; and make use of such terms as these, ’tis evident, ’tis certain, ’tis undeniable; which a due deference to the public ought, perhaps, to prevent . . . . such expressions were extorted from me . . . , and imply no dogmatical spirit, nor conceited idea of my own judgement, which are sentiments that I am sensible can become no body, and a sceptic still less than any other. But a sceptical disposition will at least put philosophical beliefs on an equal footing with ‘unphilosophical’ ones, where they have a greater chance of winning assent after due consideration. Seen in this light, Hume’s account of the ‘unphilosophical’ influences we are subjected to and of the ultimate lack of any foundation even for ‘philosophical’ belief constitutes the best sceptical lesson, and thereby the best lesson in logic, that anyone could have. In the Treatise Hume overplayed this result. Realizing the salutary effects that a sceptical disposition could have, he set out to blast his readers with the most extreme – but also the most strained – sceptical arguments he

126

9780826443595_Ch04_Final_txt_print.indd 126

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING could invent. Not content to raise sceptical doubts about causal inference in Treatise 1.3 / 69–179, Hume went on to raise sceptical arguments against the existence of an external world, against the existence of persisting substances, and even against the validity of demonstration and the existence of a self. It is as if he thought that his reader needed first to be driven into a deep sceptical crisis in order to be adequately prepared to undertake a properly scientific study of the foundations of morals in the passions, as taken up in Treatise 2 and 3.37 But he seems to have quickly realized that the excessive sceptical arguments had the opposite effect. By contesting received opinions too forcefully, he had only led readers to reject the Treatise as a whole. The Enquiry takes a different tack. Though it mentions reasons for scepticism both about the existence of an external world and about the validity of probable reasoning, it also discounts them both, presenting them merely as a means to inducing a properly scientific attitude. Hume seems to have realized that there is no better sceptical argument than the presentation of the ‘system’ with its consequences. The ‘system’, which diagnoses the problem, also cures it by means of that very diagnosis.

4

5

6

7

8

NOTES 1

2

3

For a critique of such reasons as Hume gave (at THN 1.3.2.2 / 73–4), see L. Falkenstein and D. Welton, ‘Humean Contiguity’, History of Philosophy Quarterly 18 (2001), pp. 279–96. Hume expressed some doubts about whether spatial contiguity and succession are always necessary for causality, but considered the question moot (THN 1.3.2.6 / 75 and 1.3.2.8 / 76). THN 1.3.14.14 / 162 might be read as suggesting that we have no idea of necessary connection. But Hume’s concern there was not to deny that we have the idea of a connection, or even of a necessary connection, but just that we have

9

the idea of what, specifically, connects causes necessarily to their effects. Combined with either an identification of objects with perceptions of objects (as at THN 1.4.2 / 187–218 or EHU 12 / 149–165) or an appeal to a conceivability criterion of metaphysical possibility (as at THN 1.4.5.5 / 233), this argument entails the independence of all objects occupying distinct locations in space and/or time, and so rules out the existence of necessary connections between causes and effects. The extent of Hume’s commitment to the metaphysical impossibility of necessary connections between causes and effects remains controversial. For discussion, see H. Beebee, Hume on Causation (London: Routledge, 2006) and her contribution to this volume, or the papers collected in R. Read and K.A. Richman (eds), The New Hume Debate (London: Routledge, 2000). For more on Hume’s view of intuition and demonstration, see D. Owen, Hume’s Reasons (Oxford: Oxford University Press, 1999), chap. 5. At THN 1.3.6.9 / 90–1 Hume stressed that in allowing for the possibility of hidden powers he was making a concession to his opponent for the sake of argument, not admitting that the possibility is a real one. The reader who has been keeping track of the references will note that I have juggled Hume’s order of presentation. This avoids the repetition of the same arguments under different headings that mars the exposition of the Treatise. Here what I have charitably described as a ‘meditative path of discovery’ takes on a rhetorical dimension. Prepare your reader to accept what you have to say by first inducing a deep sense of puzzlement. Then offer what you have to say as a solution to the puzzle and trust to the reader’s sense of relief to induce acceptance of your solution, even in the absence of supporting argument. Like necessary connection, the nature and role of reasoning in causal inference is controversial. For discussion, see P. Millican, D. Owen and D. Garrett in the symposia on Garrett’s Cognition and Commitment (Oxford: Oxford University Press, 1997) printed in Hume Studies 24 (1998), pp. 141–59, 171–94 and Philosophy and Phenomenological Research 62 (2001), pp. 191–6, 205–15.

127

9780826443595_Ch04_Final_txt_print.indd 127

2/9/2012 9:39:16 PM

HUME’S ACCOUNT OF PROBABLE REASONING 10

11

12

13

I here amplify on Hume’s actual statement. When discussing belief, Hume did not mention the spatial and temporal contiguity conditions he previously identified as involved in the causal relation. He spoke just of belief in the existence of an object, not of belief in its existence at a prior or subsequent time and at a contiguous place. This omission has momentous consequences, generating pseudo-problems, most notably at THN 1.4.2.21 / 197–8. An integrated account of the four features is presented at THN 1.3.7.7 / 628–9. THN 1.3.10.10 / 630–1 grapples with a further problem that might have exercised Hume: works of fiction focus attention and arouse passions without prompting belief. For further discussion of the problems with and prospects for including mental dispositions within the larger framework of Hume’s theory, see J. Bricke, Hume’s Philosophy of Mind (Princeton: Princeton University Press, 1980), chap. 3; S. Everson, ‘The Difference between Feeling and Thinking’, Mind 97 (1988), pp. 401–13; L. Loeb, Stability and Justification in Hume’s Treatise (Oxford: Oxford University Press, 2002), pp. 60–100; and J. Smalligan Marušić, ‘Does Hume Hold a Dispositional Account of Belief?’, Canadian Journal of Philosophy 40 (2010), pp. 155–83. For further discussion of Hume’s ambivalence about his account of belief see M. Bell, ‘Belief and Instinct in Hume’s First Enquiry’, in P. Millican (ed.), Reading Hume on Human Understanding (Oxford: Clarendon Press, 2002), pp. 175–85. Hume was later to argue that intuition and demonstration reduce to probability (THN 1.4.1 / 180–7). But even then his claim was not that we do not intuit relations between ideas or demonstrate truths in mathematics by appeal to a chain of intuitions. It was that because we sometimes have the wrong intuitions, our assurance of the results of a demonstration has to be informed by considerations of how likely it is that we are mistaken. Intuition and demonstration do not ‘reduce’ to probability in the sense of turning out to be nothing other than more vivacious conceptions of an idea. They ‘reduce’ to probability in the sense of presupposing second-order beliefs about the reliability of our intuitive judgements. My intuitive judgement that eight plus seven is fifteen has

14

15

16

17

18

19 20

21 22

no vivacity. My belief that I have correctly intuited this relation does. This view of the relation between the works is controversial. For an opposed view see P. Millican, ‘The Context, Aims and Structure of Hume’s First Enquiry’, in P. Millican (ed.), Reading Hume on Human Understanding, pp. 27–65, esp. pp. 40–8. Unlike the Treatise, the Abstract contains no explicit pronouncement to the effect that in speaking of ‘[t]he powers, by which bodies operate’, Hume was indulging common but false ways of thinking (cf. THN 1.3.6.8–10 / 90–1). In contrast, the parallel discussion in the Enquiry contains a qualification, this time occurring in a footnote, to the effect that the talk of hidden powers is ‘loose and popular’ and that a ‘more accurate explication’ of the notion would further buttress the conclusion to be drawn here. As noted earlier, the exposition of the Treatise is made more elegant by bringing this afterthought forward. Compare THN 1.3.16 / 176–9, which appeals to the abilities of animals as a further reason to accept the account of belief. As in the Treatise Hume continued, in both the Abstract and the Enquiry, to omit this important detail. See note 8 above. The interested reader is invited to consult THN-C: lxv–lxvii, which sets out the extent of Hume’s quotations from the Treatise in Enquiry 5, Part 2 (EHU 5.10–22 / 47–55) and his modifications to those passages. This is obviously intended. Objects can of course move around. But just as we do not assume that two objects are relatively immovable upon having once seen them alongside one another, so we do not assume that they are causally related upon once having seen them in succession. And just as the bare experience of a constant conjunction in time suffices to impel us to associate them independently of any further justification by appeal to secret powers producing a necessary connection, so the bare experience of a constant conjunction in space suffices to impel us to associate them independently of any further justification by appeal to the causes of their mobility or immobility.

128

9780826443595_Ch04_Final_txt_print.indd 128

2/9/2012 9:39:17 PM

HUME’S ACCOUNT OF PROBABLE REASONING 23

24

25

26

The effect of resemblance in raising religious devotion is of this sort. The icons and ceremonies enliven an antecedent belief in the past existence of people and events, the exception being that in this case the antecedent belief is grounded on testimony rather than memory. Belief is only enhanced by the experience of the icon or ceremony, not created. This is the first appearance of the important notion of correction by appeal to general rules. This attempt to distinguish causality from contiguity and identity is a failure. When one adds a specification of direction and distance to contiguity relations, and of temporal distance to identity relations, they become as restrictive as causal relations. Any given object is causally related to a huge number of others as well, if we do not consider whether the objects lie in the direction of cause or effect, or distinguish between proximate and remote causes and effects. When I speak here of a lack of uniformity in past experience, I mean a verified lack of uniformity, where careful scrutiny of the contiguous regions is unable to uncover any evidence of the existence of the inferred object, not an unverified lack of uniformity, where one fails to observe a cause or effect simply because one failed to look for it or (as in the case of historical inference) was in no position to observe it. Hume at one point grossly overstated the extent of the lack of uniformity in our experience, pretending that the turning of the head or closing of the eyes could prevent us from considering a succession between two types of objects to be perfectly constant (THN 1.4.2.21 / 197–8). This is an artifact of a mistake lamented in a number of previous notes: Hume’s persistent neglect of the point that, according to his own theory, causes and effects are not merely conceived to exist, but to exist at a certain place at a certain time. If I am habituated vividly to imagine a cause or effect in one place, and I consider myself to have turned my head to look at another place or to have closed my eyes and not be looking at all, I am not going to suppose that I have experienced a failure of my expectations. As Hume himself pointed out elsewhere, when we reason from causes to effects or effects to causes, it is always the case that the object we reason to is unperceived. Where both

27

28

29

cause and effect are present, the case is one of perception rather than causal inference (THN 1.3.2.2 / 73). The cases where one turns one’s head or closes one’s eyes are precisely the occasions on which causal inference is called for. Had Hume been right at THN 1.4.2.21 /197–8, there would be no such thing as causal inference. There would only be perceptions of regularities in the succession of causes and effects and perceptions of the failure of those regularities to occur, without any attendant instinct to form a belief on the latter occasions. The ‘system’ rules that possibility out without any need to invoke the elaborate mechanism proposed at THN 1.4.2.24–43 / 199–210 to provide for belief in the continued existence of objects when not perceived. These pronouncements seem inconsistent with THN 1.3.8.13–14 / 103–5, which declares that ‘in all the most establish’d and uniform conjunctions of causes and effects, such as those of gravity, impulse, solidity, &c. the mind never carries its view expressly to consider any past experience’, illustrating the claim with the point that a traveller who runs into a river does not consult past experience when forming the belief that walking out on the water will be followed by sinking and suffocating. The two passages can be reconciled if Hume’s point in THN 1.3.12.3 / 131 and 1.3.12.7 / 133 is taken to concern just those causal inferences involving new, rare and unusual objects and the formation of new causal laws. Our everyday inferences concerning familiar objects can be considered to proceed in accordance with habits learned in infancy. For Hume, mathematical results are matters of knowledge, not belief, where the knowledge arises from not being able to conceive things in any other way (THN 1.3.11.2 / 124, cf. 1.3.7.3 / 95). One can have knowledge without belief because belief involves attention, elevation of passion and inclination to action, and merely appreciating the impossibility of conceiving things any other way need have none of those results. For earlier versions of the account given here see F. Wilson, Hume’s Defence of Causal Inference (Toronto: University of Toronto Press, 1997), chap. 2, esp. pp. 123–40 and W.E. Morris, ‘Belief, Probability, Normativity’,

129

9780826443595_Ch04_Final_txt_print.indd 129

2/9/2012 9:39:17 PM

HUME’S ACCOUNT OF PROBABLE REASONING

30 31

32

33

34

in Saul Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Malden: Blackwell, 2006), pp. 77–94, esp. pp. 85–91. Hume denied this at THN 1.3.11.8 / 127. If we calculate probabilities in the common way, this one-third belief would correspond to a twothirds probability. The point to keep in mind is that Hume was not concerned about accounting for the mathematical probability of an outcome, but for the strength of belief in that outcome. For further commentary on this argument see C. Howson, Hume’s Problem (Oxford: Clarendon Press, 2000), p. 14 and chap. 4. For another ironic twist, see THN 1.3.13.12 / 149–50. In this I follow Beebee, Hume on Causation, pp. 71–4.

35

36

37

For notable instances see E 135–7; E 528–9; H 5.67; and NHR as a whole. For more on this, see D.F. Norton, ‘How a Sceptic May Live Scepticism’, in J.J. MacIntosh and H.A. Meynell (eds), Faith, Scepticism, and Personal Identity (Calgary: The University of Calgary Press, 1994), pp. 119–39, esp. pp. 128–32. This idea has been pursued in more detail by Loeb, Stability and Justification in Hume’s Treatise, pp. 6, 36, 215–29. However, Loeb is inclined to attribute Hume’s overstated scepticism to a ‘somewhat perverse’ (ibid., p. 16) aspect of his ‘temperament’ (ibid., p. 229) rather than an attempt to cause the reader to become more reflective and hesitant about all his or her beliefs.

130

9780826443595_Ch04_Final_txt_print.indd 130

2/9/2012 9:39:17 PM

5 CAUSATION AND NECESSARY CONNECTION Helen Beebee

1. INTRODUCTION It is difficult to say anything about Hume’s views on causation and necessary connection without making claims that are hotly disputed amongst interpreters of Hume’s work. Some interpreters take Hume to be a causal realist, while others hold that he is a regularity theorist. Some take him to hold that ‘causation’ is an irretrievably defective notion that could not possibly apply to any worldly phenomena, and some take him to be a non-cognitivist about our causal talk and thought. Some take him to hold that there is such a thing as objective necessary connection, while others take him to be a subjectivist about necessity. And so on. In this chapter, I shall strike a path through these and other interpretative controversies as follows. I begin in section 2 by sketching, in what I hope is a reasonably interpretatively neutral way, the bare bones of Hume’s account of causation, and in section 3 I discuss his famous ‘two definitions’. In section 4, I outline the three main classes of interpretative position – what I shall call the traditional, sceptical realist and projectivist interpretations – and briefly examine the main items of evidence that are normally marshalled for and against

each interpretation. Roughly speaking, these interpretations take Hume to be, respectively, a regularity theorist, a non-cognitivist, and a realist about causation. Finally, in section 5, I sum up the current state of play, which, as I see it, is something of a stand-off between the projectivist and sceptical realist interpretations, and say something about the specific problems that each interpretation needs to overcome if it is to prevail over its rival.

2. HUME’S BASIC ACCOUNT OF CAUSATION For Hume, causal thinking lies right at the heart of our conception of the world: all ‘reasonings concerning matters of fact and existence seem to be founded on the relation of Cause and Effect. By means of that relation alone we can go beyond the evidence of our memory and senses’ (EHU 4.3 / 26). In other words, our access to external reality, beyond the ‘evidence’ of current experience and memory, entirely depends on reasoning from causes to effects (and vice versa). When I form the belief that my dinner will not poison me, or that what I am reading in the newspaper

131

9780826443595_Ch05_Final_txt_print.indd 131

2/9/2012 9:34:59 PM

CAUSATION AND NECESSARY CONNECTION is true, or that the kettle I turned on a few minutes ago will have boiled by now, I do so on the basis of such reasoning. This being so, the primary task that Hume sets himself is to uncover the nature of this mysterious relation – or rather, to discover what our idea of causation consists in, since it is ‘impossible to reason justly, without understanding perfectly the idea concerning which we reason; and ’tis impossible to understand any idea, without tracing it up to its origin, and examining that primary impression from which it arises’ (THN 1.3.2.4 / 74–5). He quickly discovers that ‘whatever objects are consider’d as causes and effects are contiguous’ (THN 1.3.2.6 / 75) and are also such that the cause is temporally prior to its effect. But those two conditions clearly do not exhaust our idea of causation, for ‘there is necessary connexion to be taken into consideration’ (THN 1.3.2.11 / 77). Hume’s search for the impression-source of the idea of necessary connection is a long one, for it turns out that this crucial component of ‘the idea concerning which we reason’ has its source in that very reasoning itself. So Hume needs to discover the nature of our reasoning from causes to effects (what I shall call ‘causal reasoning’) before he can locate the impression-source of the idea of necessary connection. Hume’s investigation into causal reasoning – what is traditionally described as his discussion of the ‘problem of induction’ – yields two results that are significant for present purposes. First, causal reasoning proceeds on the basis of past observed regularity: on observing a C, we infer that an E will follow just when we have experienced sufficiently many Cs followed by Es. Second, that reasoning proceeds not by consideration of any argument, but as a matter of ‘custom or habit’ (EHU 5.5 / 43). There

is simply a mental mechanism that, given relevant past experience, conveys the mind from the impression of a C (one billiard ball striking another, for example) to a belief that an E will follow (the second ball moving). And it turns out, in ‘Of the idea of necessary connexion’ (THN 1.3.14 / 155–72; EHU 7.26–30 / 73–9), that it is the operation of this very mechanism that furnishes us with the impression, and so the idea, of necessary connection. Before Hume can establish this latter claim, however, he needs to show that we have no sensory impression of necessary connection. Before briefly rehearsing his argument, it is worth saying something about the importance of the issue for Hume. One of Hume’s main aims is to provide a ‘science of man’: an account of how the mind works that is based on a clear-headed, ‘experimental’ investigation. A plausible account of his primary target is given by Edward Craig,1 who sees Hume’s major adversaries as those who uphold what he calls the ‘Image of God’ doctrine (we are made in God’s image) and a corresponding epistemological doctrine: the ‘Insight Ideal’. According to the Insight Ideal, the nature of reality – or at least some of it – is in principle accessible to reason; thus philosophers before Hume had subscribed to the self-evident or a priori status of the claim that every event has a cause, or had claimed that the essence of objects can be known by what Descartes calls ‘purely mental scrutiny’.2 Hume, by contrast, sets out to undermine systematically the claim that any aspect of the nature of reality is knowable a priori, and, moreover, to show that no ‘matter of fact’ can be inferred a priori from any other distinct matter of fact. Hume takes himself to have established this claim in his discussion of causal reasoning; but it is a claim to which he returns in the negative phase of his discussion of the

132

9780826443595_Ch05_Final_txt_print.indd 132

2/9/2012 9:35:00 PM

CAUSATION AND NECESSARY CONNECTION idea of necessary connection (EHU 7.1–25 / 60–73; THN 1.3.14.1–18 / 155–65), where he argues that the idea does not have a sensory impression-source. Hume assumes, in this discussion, that such an impressionsource for the idea of necessary connection would have to be such that it delivers certainty that the effect will follow: ‘were the power or energy of any cause discoverable by the mind, we could foresee the effect, even without experience [of past constant conjunction]; and might at first, pronounce with certainty concerning it, by the mere dint of thought and reasoning’ (EHU 7.6 / 63). With this assumption (to which I return below) in place, it is an easy matter to establish Hume’s negative conclusion, since, as we already know from his discussion of causal reasoning, observation of a particular event never delivers such certainty: we can always imagine the cause happening without its effect, and so it is always epistemically possible that the effect will not occur. The assumption just mentioned has caused much puzzlement amongst commentators. Hume is apparently arguing for a phenomenological claim – that on first observing them, ‘[a]ll events seem entirely loose and separate. One event follows another; but we can never observe any tie between them’ (EHU 7.26 / 74) – and yet his argument for this claim proceeds by way of pointing out that we cannot ‘pronounce with certainty’ that the effect will follow, on observing the cause. But how is the former claim supposed to follow from the latter? In particular, is Hume not confusing two distinct notions? On the one hand, we have the claim that there is no observable power, within the cause itself (e.g. the striking of one billiard ball by another), such that observing it would deliver certainty that the effect will follow. (J.L. Mackie calls such a power ‘necessity2’.3) But this does not

entail that there is no observable connection or ‘tie’ between cause and effect (necessity1). It might easily be that we can observe such a connection – that is, we can observe the causal relation – despite the fact that we cannot observe any power, in the cause itself, that produces certainty about the effect. It seems that Hume does indeed run these two distinct notions together; however, by his own lights it is not clear how much of a problem this is. Hume’s central concern, remember, is with inference from one matter of fact to another – that is, from causes to effects (and vice versa). The observability of necessity1 – of a mere causal ‘tie’ between causes and effects – would make no substantive difference to the account that Hume has already offered of such reasoning, since such a tie could only be observed by, as it were, observing c-causing-e as a package deal, and could therefore not serve as the basis of causal reasoning. For how would we then reason, when confronted with a C? The best we could do is reason that, since Cs have always been observed to cause Es in our past experience, the currently observed C will likewise cause an E. But this inference cannot be a priori, since it is still epistemically possible that the former is true and the latter false. So we need to postulate a different mental mechanism that supplies the inference, and that mechanism would turn out (according to Hume’s own argument) to be custom or habit. Thus the only difference between the account just canvassed and Hume’s own account of causal reasoning is that according to the former account the impression of causation is supplied by sensation, whereas according to Hume’s account it has another source. That source, of course, turns out to be the very inference from causes to effects itself: the ‘transition arising from the accustom’d

133

9780826443595_Ch05_Final_txt_print.indd 133

2/9/2012 9:35:00 PM

CAUSATION AND NECESSARY CONNECTION union’ (THN 1.3.14.19 / 165), that is, the habit that takes the mind from an impression of the cause, together with experience of past constant conjunction, to the belief that the effect will follow. As Hume puts it, when one particular species of event has always . . . been conjoined with another, we make no longer any scruple of foretelling one upon the appearance of the other, and of employing that reasoning, which can alone assure us of any matter of fact or existence. We then call the one object, Cause; the other, Effect. (EHU 7.27 / 74–5) Hence the ‘connexion . . . which we feel in the mind, this customary transition of the imagination from one object to its usual attendant, is the sentiment or impression from which we form the idea of power or necessary connexion’ (EHU 7.28 / 75). Hume thus finally achieves what he set out – nearly a hundred pages earlier, in the Treatise (THN 1.3.2.4 / 74) – to achieve: the impression-source for the idea of necessary connection. Unfortunately, however, it is far from clear what consequences Hume takes his discovery to have for our causal talk and thought, for he appears to endorse three positions that are mutually inconsistent: 1. He seems to think that we are apt to project the impression – and the idea – of necessary connection into the world: ‘the mind has a great propensity to spread itself on external objects, and to conjoin with them any internal impressions, which they occasion’ (THN 1.3.14.23 / 167). And he seems to suggest that this projection is a mistake: ‘we are led astray by a false philosophy’ when ‘we transfer the determination of the thought to external objects, and suppose any real intelligible connexion betwixt them; that being a quality, which can only belong to the

mind that considers them’ (THN 1.3.14.25 / 168). This suggests that it is a mistake to think that events in the world really are necessarily connected to one another. 2. He does not appear to suggest that the idea of causation can be stripped of the component idea of necessary connection: he does not appear to respond to the mistake just identified by advocating a revisionary account of the concept of cause. 3. He appears to think that our causal thought and talk is truth-apt. (Certainly it is subject to normative constraints: in the Treatise the very next section lists ‘rules by which to judge of causes and effects’ (THN 1.3.15 / 173–6).) The inconsistency is easy to see: by (2), we really do deploy the idea of necessary connection when we engage in causal talk and thought; and, by (3), that causal talk is in general entirely legitimate (not least because it is subject to normative constraints; it is hard to see how this could be so if all such talk was irredeemably defective). And yet, by (1), that talk is irredeemably defective. There can be nothing in the world that answers to the idea of necessary connection, since that idea derives solely from the ‘determination of the mind’. The three broad interpretative rivals discussed in section 4 below resolve the inconsistency in different ways. Roughly speaking, the traditional interpretation denies (2),4 and both the projectivist and sceptical realist interpretations finesse (1): they hold Hume to the claim that there is a mistake in the offing, but deny that the mistake in question is that of thinking of events in the world as causally or necessarily connected. Finally – and this is what distinguishes these two interpretations – the projectivist interpretation finesses (3) as well: Hume does endorse our causal talk and thought, but that talk and thought is to be understood in

134

9780826443595_Ch05_Final_txt_print.indd 134

2/9/2012 9:35:01 PM

CAUSATION AND NECESSARY CONNECTION non-cognitivist terms. For the sceptical realist, by contrast, our causal talk and thought is straightforwardly referential: there is something in the world that answers to our idea of causation (or at least we believe that there is, and there is nothing defective about that belief). One more piece of the already difficult puzzle needs to be put on the table, namely Hume’s famous two definitions of causation; this is the topic of the next section. Before leaving Hume’s discussion of the origin of the idea of necessary connection, however, it is worth noting what is, in my view, an often misunderstood feature of his account. It is routinely taken for granted that in Hume’s view, all events ‘seem entirely loose and separate’. That is, phenomenologically speaking, our experience is merely as of one event following another, even once the habit of inference has been established, and so even once the impression of necessary connection is present. This has forced some interpreters to cast the impression of necessary connection as, for example, simply a ‘peculiar feeling’5 or a ‘feeling of helplessness or inevitability’(where the inevitability is the inevitability of one’s expectation, and not the inevitability of the effect itself).6 Hume is not, in fact, committed to this view. What he says is that ‘all events seem entirely loose and separate’; he does so in the context of ‘single instances of the operation of bodies’ (EHU 7.26 / 73) – that is, when we first observe a pair of contiguous events, and hence before the habit of association has arisen. He does not explicitly state a view about how things seem once the habit of association has arisen, but it is plausible to suppose that he takes the impression of necessary connection to affect, precisely, how things seem – that is, to affect the nature of visual experience itself.

This interpretation removes the need to think of the impression of necessary connection as a ‘feeling of expectation’ or similar; how things seem, when the impression arises, is, precisely, necessarily connected, just as it is by virtue of the impression of red that things seem red to us. It also helps to explain why Hume offers a non-phenomenological argument for the claim that there is no sensory impression of necessary connection. After all, if all events really did seem entirely loose and separate to us, Hume would not need such an argument; unbiased phenomenological reflection would do the job just by itself (and would thereby establish the stronger claim that we have no impression of either necessity1 or necessity2).7

3. THE TWO DEFINITIONS Here are Hume’s two definitions of causation – or, more precisely, definitions of ‘a cause’ – as they appear in the Treatise, towards the end of his discussion of the idea of necessary connection: (D1) ‘An object precedent and contiguous to another, and where all the objects resembling the former are plac’d in like relations of precedency and contiguity to those objects, that resemble the latter’. (THN 1.3.14.29 / 170) (D2) ‘[A]n object precedent and contiguous to another, and so united with it, that the idea of the one determines the mind to form the idea of the other, and the impression of the one to form a more lively idea of [that is, a belief in] the other’. (ibid.) Unfortunately, rather than clarifying the situation, the two definitions are problematic in

135

9780826443595_Ch05_Final_txt_print.indd 135

2/9/2012 9:35:01 PM

CAUSATION AND NECESSARY CONNECTION their own right. The first thing to note is that they are not even extensionally equivalent. The conditions in (D2) can easily be satisfied without (D1) holding, by someone having observed an unrepresentative sample of ‘objects’ (or events), so that the two kinds of event are constantly conjoined in their experience but not universally. And (D1) can be satisfied without (D2) holding, for instance in the case of constant conjunction between two kinds of events that nobody has observed, so that nobody’s mind is determined to move from the idea of one to the idea of the other. A standard solution to this problem8 has been to claim that Hume only intends (D1) as a genuine definition of causation, while (D2)’s aim is different: to explain the conditions under which we do, in fact, come to make causal claims, for example. This move seems somewhat ad hoc, however, given Hume’s claim that the two definitions present ‘a different view of the same object’ (THN 1.3.14.29 / 170). A second solution, offered by Don Garrett,9 is to distinguish a ‘subjective’ from an ‘absolute’ reading of each definition. Roughly, a subjective reading of (D1) would take ‘all objects’ to mean all objects observed so far by a particular person. This would then be coextensive with (D2) read subjectively, with ‘the mind’ understood as referring to the same person, since an ‘object’ that meets (D1) now will be such that the mind of the person in question is indeed determined by the idea of the object to form the idea of its effect. As Garrett puts it, read subjectively, the two definitions tell us when an object ‘functions psychologically’ as a cause.10 An absolute reading of (D1) takes ‘all objects’ to be unrestricted, so that (D1) appeals to universal constant conjunction. This is coextensive with (D2) read absolutely, with ‘the mind’ now understood to refer to some sort

of ‘idealized spectator’, that is, an observer who observes representative samples of all kinds of constantly conjoined events. A third solution – similar to Garrett’s ‘subjective’ reading and to Robinson’s proposal concerning the second definition – is to read both definitions not as ‘definitions’ in the standard contemporary sense at all, but as saying how it is we come to believe that one thing is a cause of another. As Edward Craig puts it, the definitions characterize the ‘circumstances under which belief in a causal connection arises, one concentrating on the outward situation, the other on the state of the believer’s mind that those outward facts induce’.11 My own view is that none of these solutions are satisfactory, because they all ignore Hume’s preceding remark in the Treatise that ‘two definitions of this relation may be given of this relation, which are only different, by their presenting a different view of the same object, and making us consider it either as a philosophical or as a natural relation; either as a comparison of two ideas, or as an association betwixt them’ (THN 1.3.14.31 / 169–70). Hume’s distinction between philosophical and natural relations is a distinction between two kinds of mental operation. Roughly, the former is the conscious ‘placing’ of two ideas under a relation (hence ‘plac’d’ in (D1)), and the latter is the unconscious ‘transition’ of the mind from one idea to another. For example, resemblance – which is the other relation that is both natural and philosophical – can operate in two distinct ways, as when I consider whether a painting of a particular scene resembles the image of a particular remembered scene I have in my mind and come to judge that it does (philosophical relation), or when I see a picture of the Queen and my mind is automatically drawn to the idea of the Queen herself (natural relation). Similarly

136

9780826443595_Ch05_Final_txt_print.indd 136

2/9/2012 9:35:01 PM

CAUSATION AND NECESSARY CONNECTION for causation: I can ‘place’ two ideas under the relation of causation, and will (or should) do so precisely when the conditions specified in (D1) are met (contiguity, precedence and observed constant conjunction), thereby coming to form the judgement that one event is the cause of the other. And I ‘naturally’ judge one event to be the cause of another when I have acquired the relevant habit of association: my mind is drawn from the idea (or impression) of the first event to the idea of (or belief in) the second, and, again, I thereby come to form the judgement that the first event is the cause of the second.12 Note that neither of the last two of the four interpretative positions just outlined delivers any verdict about the meaning of ‘cause’, since both deny that the definitions are definitions in anything like the standard contemporary sense. Instead, the definitions tell us something about how it is we come to make causal judgements. As we shall see, the availability of these interpretative options with respect to the two definitions undermines a key component of the motivation for the traditional interpretation of Hume on causation, according to which he is a naive regularity theorist. We are not required to read the first definition as a definition of the meaning of ‘cause’, and so we are not required (at least not required by the two definitions) to hold Hume to the view that causation consists in contiguity, precedence and constant conjunction.

4. INTERPRETATIONS: TRADITIONAL, SCEPTICAL REALIST AND PROJECTIVIST It is uncontroversial that Hume endorses the following three theses. First, at least

in the most basic cases, we come to make causal judgements – judgements of the form ‘c caused e’, or ‘c will cause e’ – on the basis of the temporal priority and contiguity relations that hold between these two events, and the observed constant conjunction between events of the kinds that c and e instantiate. In such cases we infer that e will occur on the basis of having observed c, and we also think of c as a cause of e, in a way that, somehow or other, involves deploying the idea of necessary connection. Second, the impression-source of that idea is the inference just described: the impression of necessary connection is not a sensory impression but an impression of reflection. Finally, there is – or can be – something awry in our deployment of the idea of necessary connection. Perhaps the easiest way of seeing the difference between the three main interpretative positions is to consider how they interpret the second and third of the claims just described. I shall thus start my brief account of each of the positions by describing their attitudes to those claims. THE TRADITIONAL INTERPRETATION According to what I am calling the ‘traditional’ interpretation, Hume is a naive regularity theorist about causation: c causes e if and only if c is contiguous with and prior to e, and events similar to c (the Cs) are constantly conjoined with events similar to e (the Es).13 In other words, ‘c causes e’ just means ‘all Cs are followed by Es’. What is awry in our deployment of the idea of necessary connection, in this view, is, precisely, that we deploy it at all: given that the source of that idea is the transition of the mind, it cannot possibly represent any mind-independent feature of the world. Thus (and this is an issue on which versions of the traditional interpretation differ) either (i) necessary connection is

137

9780826443595_Ch05_Final_txt_print.indd 137

2/9/2012 9:35:01 PM

CAUSATION AND NECESSARY CONNECTION not, in fact, part of the meaning of ‘cause’ as we actually deploy the idea ‘cause’, or (ii) it is part of the ordinary meaning of ‘cause’ but Hume is in effect offering a revisionary conceptual analysis of ‘cause’ (shorn of the troublesome concept of necessary connection), or (iii) it is part of the ordinary meaning of ‘cause’ and Hume offers no such revision: ‘cause’ is irredeemably defective. The major problem with the traditional interpretation is an almost total lack of evidence in its favour. The major source of evidence has standardly been thought to be the ‘two definitions’ – or rather, the first definition. However, as we saw in section 3, the interpretation of the two definitions relied on here (according to which the first definition really is Hume’s attempt to offer a conceptual analysis of ‘cause’) has recently come under attack. In addition, the accounts given of the role of the idea of necessary connection offered by different versions of the traditional interpretation fit badly with the text. Against alternative (i) above, Hume does not suggest (setting aside the first definition) that the idea of necessary connection is no part of the actual meaning of ‘cause’; nor, contra (ii), does he suggest that the concept of causation needs to be revised. He does, of course, suggest that something is apt to go awry when we deploy the idea of necessary connection; but he does not suggest that the appropriate response is to stop deploying it; indeed, it is hard to see how Hume could even consider this to be a genuine psychological possibility. Finally, against (iii), which I take to be Barry Stroud’s position, Hume clearly and persistently endorses a wide range of causal claims, and indeed provides ‘rules by which to judge of causes and effects’ (THN 1.3.15 / 173–6). This is extremely difficult to square with the claim that he takes all causal talk to be equally and irredeemably false.

SCEPTICAL REALIST INTERPRETATIONS In stark contrast to the traditional interpretation, the sceptical realist interpretation takes Hume to hold that our causal talk refers – successfully – to real, mind-independent causal connections in nature, or what I shall call ‘causal powers’.14 Exactly what role the idea of necessary connection plays here is something that sceptical realist interpreters differ on. In particular, John Wright holds that the idea of necessary connection refers to genuine mind-independent necessary connections, so that if only we could penetrate into the true, underlying nature of causes, we would be able to discern those (causal) powers by which causes really do absolutely necessitate their effects: he ‘retained an ideal of knowledge of true causes which was derived from the Cartesians’,15 so that the ‘true manner of conceiving a particular power in a particular body’ (of which we are in fact incapable) would involve being ‘able to pronounce, from a simple view of the one, that it must be follow’d or preceded by the other’ (THN 1.3.14.12 / 161).16 Thus the mistake we are apt to make when it comes to the idea of necessary connection is the mistake of holding that it derives from an impression of sensation, and thus holding that we really do perceive necessity. Hence our idea of necessary connection is defective, in that – being derived from an impression of reflection – it cannot adequately represent real necessity in nature; nonetheless, it succeeds in referring to real necessity. In other words, Hume agrees with his opponents when it comes to the metaphysics of the Image of God doctrine, as far as causation is concerned, but he disagrees with them over the Insight Ideal: we lack the God-like ability to penetrate into the essences of things in a way that would reveal their true, effect-guaranteeing, underlying nature to us.

138

9780826443595_Ch05_Final_txt_print.indd 138

2/9/2012 9:35:02 PM

CAUSATION AND NECESSARY CONNECTION According to Galen Strawson, by contrast, Hume takes all necessity (whether causal or logical) to be purely subjective: our idea of necessity does not track any real necessity in nature, but is ‘just a feeling we have about certain things – about 2 + 2 = 4, and about what this billiard ball does to that one, and about the sum of the angles of a triangle’.17 Nonetheless, our idea of causation succeeds in referring to more in the world than mere regularities: it refers to (to use Hume’s own expression) that upon which the ‘regular course and succession of objects totally depends’ (EHU 5.22 / 55); or, in Strawson’s words, ‘whatever it is about the universe (or matter) which is that in virtue of which it is regular’.18 Again, a central thought here is that our idea of causation is inadequate to what it represents: we can have a ‘relative’ idea of it, but not what Strawson calls a ‘positively contentful’ idea of it. Thus Wright’s Hume conceives of the referent of ‘cause’ to be objective necessity – the feature of the cause that absolutely guarantees that the effect will occur – whereas Strawson’s Hume conceives of the referent of ‘cause’ to be a regularity-guaranteeing feature of nature. The starting point for the claim that there is serious textual evidence to justify (some version of) the sceptical realist interpretation is to take the first Enquiry rather than the Treatise as expressing Hume’s considered view about causation. This is because it is in the first Enquiry that Hume refers, on several occasions, to the ‘powers and forces, by which [the course of nature] is governed’ that are ‘wholly unknown to us’ (EHU 5.21 / 54), the ‘powers and principles on which the influence of these objects entirely depends’ that nature ‘conceals from us’ (EHU 4.16 / 33), and so on. It is also the place where he says that our idea of causation is ‘imperfect’, and that it admits of no ‘just definition,

except what is drawn from something extraneous and foreign to it’ because we ‘have no idea of this connexion, nor even any distinct notion what it is we desire to know, when we endeavour at a conception of it’ (EHU 7.29 / 76–7). It is possible to reinterpret these claims in a way that does not commit Hume to belief in causal powers. In particular, one might take the claims about nature’s ‘secret’ powers to be mere suppositions for the sake of the argument.19 Or one might point out that, since Hume explicitly takes ‘power’ to be synonymous with ‘cause’ (THN 1.3.14.4 / 157), and since he doubtless thinks that there are additional, not-yet-known regularities underlying the observed behaviour of objects, talk of secret powers presents no problem for the other interpretations.20 But there is nothing in the text of the first Enquiry itself to motivate such a reinterpretation (or so defenders of sceptical realism maintain): the most natural interpretation of the first Enquiry, taken on its own terms, reads Hume’s claims at face value, as expressions of belief in, but ignorance concerning the nature of, causal powers. One advantage of Wright’s version of the sceptical realist interpretation over the traditional interpretation is that it makes sense of the tension noted in section 2: Hume endorses our (necessary-connectioninvolving) causal thought and talk, and yet he apparently thinks there is something wrong with the idea of necessary connection. Wright’s account resolves the tension by identifying what is ‘wrong’ with the idea of necessary connection as our tendency to think that necessary connections are perceivable, and our corresponding tendency to think that we have thus penetrated into the essence of bodies: ‘[t]he vulgar mistake an associational connection for a genuinely perceived

139

9780826443595_Ch05_Final_txt_print.indd 139

2/9/2012 9:35:02 PM

CAUSATION AND NECESSARY CONNECTION rational connection’.21 This tendency does not, according to Wright’s account, affect the meaning of our causal talk, however. Our habits of inference give rise to the belief that there are necessary connections in nature, and that belief is (for all we know) true; the mistake we tend to make is the mistake of thinking that our idea of necessary connection is fully adequate to what it represents. When we ‘make the terms of power and efficacy signify something, of which we have a clear idea, and which is incompatible with those objects, to which we apply it, obscurity and error begin then to take place’ (THN 1.3.14.25 / 168). This is because we are in effect claiming that there is some feature of the external world that is adequately represented by an idea whose impression-source is a mere transition in the mind, when in fact there can be no such feature. But this mistake plays no role in the meaning of ‘cause’: our idea of necessary connection really does refer to real necessity, even if we are apt to be mistaken about the nature of what it is we thereby represent. It is unclear whether Strawson’s interpretation also has this advantage over the traditional interpretation, given his claim that Hume takes all necessity to be subjective. Strawson says that ‘the E-intelligible [that is, positively contentful] meaning of the term “causation” can only encompass certain aspects of the experience Causation [that is, causal power] gives rise to . . . [including] the feeling of determination’.22 So according to Strawson’s account, the idea of necessary connection is only a part of the ‘meaning’ of ‘causation’ in the sense that it is the idea of an experience that causation ‘gives rise to’. But of course such a claim could equally be made by a defender of the traditional interpretation; so if Strawson’s claim here makes adequate sense of the fact that Hume

endorses our necessary-connection-involving causal talk and thought, the same can be said of the traditional interpretation. My own view is that Strawson’s suggestion does not adequately capture the thought that the idea of necessary connection really is part of the meaning of ‘cause’: Hume really does seem to think that our causal thought involves the claim that causes and effects are necessarily connected, and not merely that they happen to give rise to a certain kind of ‘feeling’. A second advantage that the sceptical realist interpretation has been claimed by Strawson to have over the traditional interpretation is that the latter saddles Hume with the preposterous (according to Strawson) claim that there is nothing more to the world than mere regularity (‘one of the most baroque metaphysical suggestions ever put forward’).23 The traditional interpretation need make no such claim, however. According to the traditional interpretation, Hume’s claim is only that our thoughts cannot successfully reach out to any mind-independent relations between causes and effects, aside from priority and contiguity. This is not to assert positively that such relations do not, or cannot, exist – only that we cannot succeed in referring to them in our causal thought and talk (if we try, we ‘lapse into obscurity and error’). We can (as Strawson says) form a ‘relative’ idea of such relations – we can consider the possibility that they exist (without being able to form adequate ideas of what they might be like), but to do so would be, at best, to indulge in idle metaphysical speculation. From a semantic point of view, then, the crux of the difference between the traditional interpretation and Strawson’s version of sceptical realism is that for Strawson, the ‘relative’ idea we form is a relative idea of real causal powers, which are what our ordinary causal talk refers to. According to the traditional interpretation,

140

9780826443595_Ch05_Final_txt_print.indd 140

2/9/2012 9:35:03 PM

CAUSATION AND NECESSARY CONNECTION by contrast, we can form a relative idea of some possible not-further-specifiable relation between causes and effects, but that idea is not the idea of causation: the idea of causation is exhausted by contiguity, priority and constant conjunction. THE PROJECTIVIST INTERPRETATION24 The projectivist interpretation in some sense represents a middle ground between the sceptical realist and traditional interpretations (indeed Angela Coventry calls it the ‘intermediate interpretation’).25 It shares with sceptical realism the thought that our causal talk and thought does more than merely register the existence of regularities in nature, and with Wright’s version of sceptical realism the thought that this involves the legitimate deployment of the idea of necessary connection. However, the projectivist interpretation shares with the traditional interpretation a broadly meaning-empiricist interpretation of Hume, according to which experience places strict limits on what can be represented, via our ideas, in our thought and talk. In particular, the two interpretations agree that nothing in reality answers to our idea of necessary connection: causation ‘in the objects’, as it is sometimes put, amounts to no more than contiguity, priority and constant conjunction. The projectivist interpretation squares the apparent circle by adopting a non-cognitivist approach – on the one hand, the idea of necessary connection is deployed legitimately in our causal talk and thought, but on the other, no aspect of reality answers to this idea. A standard interpretation of Hume’s ethical (and indeed aesthetic) writing takes him to be a ‘sentimentalist’ about moral (and aesthetic) claims: when we make an evaluative claim about an action or a person (that they are good or bad, brave or cowardly,

and so on), we are expressing a sentiment or moral attitude towards that action or person, rather than attributing to them some mindindependent moral property. Moreover (this further move is admittedly more controversial), we do not merely express the relevant sentiment; we project it onto the object of our experience or judgement, so that the painting we are observing looks beautiful or the murder will seem vicious thanks to the projection of the relevant sentiment, and we correspondingly judge them to be so thanks to the projection of the relevant idea. Thus Michael Smith notes that when Hume says that you ‘never can find’ the viciousness of a murder ‘till you turn your reflexion into your own breast, and find a sentiment of disapprobation’ (THN 3.1.1.24 / 468–9), he ‘is precisely trying to focus our attention away from where it is naturally focused when we judge a wilful murder to be wrong: that is, away from the murder itself, and onto an otherwise quite unnoticed “calm passion” he supposes to arise in us’.26 Part of the point here, according to the projectivist line on morality, is that Hume is not merely making a straightforward phenomenological claim; it is not supposed to be just obvious to us that there is nothing in the murder itself that constitutes its viciousness. On the contrary, he only gets to this claim after quite a lot of argument. So – the thought is – it seems to us that our moral and aesthetic judgements are responses to genuine features of external objects, and their so seeming is due to the projection of the relevant sentiment. Our sentiment-derived judgements are, however, subject to normative constraints – so, for example, there are ‘rules of art’ which deliver a true standard of taste and sentiment’,27 even though ‘no sentiment represents what is really in the object’.28 Hence moral and aesthetic claims

141

9780826443595_Ch05_Final_txt_print.indd 141

2/9/2012 9:35:03 PM

CAUSATION AND NECESSARY CONNECTION can legitimately be regarded as correct or incorrect (there is a ‘true standard’ for them to meet) on the basis of those normative constraints.29 In the case of causation, the analogue of ‘sentiment’ is, of course, the impression – and hence the idea – of necessary connection. According to the projectivist view, we are apt to mistake the impression of necessary connection for a sensory impression (this is a point of agreement with Wright), and we are apt to do this because we project the impression onto the external objects that trigger it. (My own proposal here is that the ‘impression’ of necessary connection is in fact, for Hume, not a self-standing ‘feeling’ at all, but merely the modification of visual experience that occurs once our habit of expectation is formed; see section 2 above.) And when we ‘describe’ what is going on – when we ‘call the one object, Cause; the other, Effect’ (EHU 7.27 / 74–5) – we similarly project the idea of necessary connection onto those objects. Our causal talk and thought, then, is not descriptive: it does not attribute a mind-independent relation to causes and effects, but projects our idea of necessity onto them.30 As with the moral and aesthetic cases, however, this is not to say that Hume is a subjectivist about causation, for there are norms that govern the appropriateness of our causal claims to their objects, for example in his ‘rules by which to judge of causes and effects’ (THN 1.3.15 / 173–6).31 More generally, Hume is certainly in a position to regard our natural, instinctive causal judgements as eminently revisable in the light of the evidence.32 Thus, for example, As might all have been followed by Bs in my experience, so that I judge that the As are causes of Bs. But I might then find out that there are two distinct kinds of A (call them A1s and A2s), that all the As I have observed are in

fact A1s, and that A2s are often not followed by Bs. This would give me good grounds to revise my initial judgement that As cause Bs in favour of the judgement that A1s cause Bs but A2s do not, since the ‘difference in the effects of two resembling objects must proceed from that particular, in which they differ’ (THN 1.3.15.8 / 174; rule 6). And this would be so even if the habit of inference that has been established naturally inclines me, on next observing an A, to expect a B, and thus to judge that they are causally connected if a B does indeed follow; I know that that judgement is hostage to information I do not possess, namely whether or not the observed A is an A1 or an A2. In other words, the relevant rule acts as a normative constraint on the causal judgement I am naturally inclined to make. Direct textual evidence for the projectivist interpretation is admittedly rather thin. Indirect support comes largely from similarities between Hume’s treatment of causation on the one hand, and moral and aesthetic judgements on the other. In each case, we have the thought that our judgement (causal, moral or aesthetic) ‘adds’ something to mindindependent matters of fact. In the causal case, Hume says that the mind does this via its ‘propensity to spread itself on external objects’ (THN 1.3.14.23 / 167); in the moral and aesthetic cases, he postulates ‘a productive faculty, and gilding or staining all natural objects with the colours, borrowed from internal sentiment, raises in a manner a new creation’ (EPM App. 1.21 / 294). A further piece of evidence – and at the same time a response to the charge that Hume shows no serious positive inclination towards a non-cognitivist account of causation – comes from his selective use of the terms ‘matter of fact’ and ‘belief’. Hume never talks about causal ‘beliefs’ or considers

142

9780826443595_Ch05_Final_txt_print.indd 142

2/9/2012 9:35:03 PM

CAUSATION AND NECESSARY CONNECTION causal claims to fall within the class of ‘matters of fact’. We make causal judgements, but these would appear not to have the status of belief for Hume, nor would they appear to be judgements about matters of fact – or at least, Hume never claims that they are.33 In other words, Hume appears to restrict ‘matter of fact’ – and correspondingly, ‘belief’ – to the relata of causation. From the perspective of the traditional and sceptical realist interpretations, this is rather puzzling. For in both interpretations, there are perfectly good facts about causation to be had, and so our causal ‘judgements’ should count as beliefs, every bit as much as the existence of a moving billiard ball is a matter of fact and our expectation that the billiard ball will move is a belief. From the perspective of the projectivist interpretation, by contrast, there is no real anomaly here. Causal reasoning, Hume says, is reasoning from one matter of fact to another. If causal judgements are projections of the idea to which this reasoning gives rise, then those judgements are not beliefs about matters of fact, any more than moral and aesthetic judgements are beliefs about matters of fact: they are projections of our habits of thought onto matters of fact, and so do not constitute beliefs about matters of fact.34

5. PROBLEMS AND PROSPECTS For quite a large part of the twentieth century, most analytic philosophers steadfastly avoided appealing to the concept of cause: a ‘horrid little word’, according to Peter van Inwagen,35 and ‘a truly obscure’ concept, according to John Earman.36 In particular, discussion of the problem of induction overwhelmingly proceeded as though inference from the observed to the unobserved

is largely independent of beliefs about the causal structure of reality. (For example ‘causation’ does not even appear in the index of Colin Howson’s 2000 book, Hume’s Problem: Induction and the Justification of Belief [Oxford: Oxford University Press].) Hume is, of course, the philosopher from whom we are supposed to have learned that causal thinking is both suspect and dispensable. The irony is that this is not Hume’s view at all: for Hume, causal thinking is central to our understanding of, and beliefs about, the nature of reality. But how are we to understand what ‘causal thinking’ amounts to for Hume? As we have seen, the range of interpretative possibilities is very wide indeed. My own view is that the traditional interpretation, in all its forms, is untenable: Hume holds that causal thinking amounts to more than belief in regularities, for it involves the idea of necessary connection. Moreover, that idea is entirely legitimate, when correctly understood, and it does serious philosophical work for him.37 But this leaves both projectivism and at least one version of sceptical realism still in the running, and deciding between these possibilities is a difficult task. Part of the difficulty lies in the differences between the Treatise and the first Enquiry. Reading the first Enquiry as a reworking of the Treatise, with no substantial change in the philosophical views presented, inclines one towards projectivism, correspondingly encouraging one to reinterpret Hume’s talk of secret powers and the like in the Enquiry so that they do not express a commitment to the existence of real causal powers. Reading the first Enquiry on its own, by contrast, with no preconceptions carried over from reading the Treatise, on the grounds that Hume took the first Enquiry to best express his considered philosophical view, inclines one towards sceptical realism. But of course which strategy

143

9780826443595_Ch05_Final_txt_print.indd 143

2/9/2012 9:35:03 PM

CAUSATION AND NECESSARY CONNECTION one should adopt is largely a question of historical fact that further attention to the texts themselves will not resolve.38 There is, however, at least one reason to be sceptical about sceptical realism. Sceptical realism (or at least Wright’s and Kail’s versions thereof) inevitably saddles Hume with a deeply puzzling view. On the one hand, all sides agree that Hume rejects what Craig calls the ‘Insight Ideal’: we cannot penetrate into the ‘essence’ of bodies in such a way as to reveal any features that would license a priori inferences from causes to effects. Nonetheless, Wright’s and Kail’s versions of sceptical realism attribute to Hume one aspect of the ‘Image of God’ doctrine, the metaphysical position that underlies the Insight Ideal, for they both attribute to Hume ‘an ideal of knowledge of true causes which was derived from the Cartesians’.39 In other words, ‘true causes’ are such that we would, if only we could penetrate into their nature, be able to infer effects from causes a priori. And the question is: why would Hume commit himself to such a metaphysical position? After all, the only motivation for such a view is the thought that the nature of reality must be such that God himself (and so we, if we are sufficiently God-like) can infer effects from causes a priori. And this is a motivation that Hume himself clearly lacks: he has no reason whatsoever to want to cling on to a picture of the nature of reality that derives from views about God’s epistemic access to that reality.

NOTES 1

2

E. Craig, The Mind of God and the Works of Man (Oxford: Clarendon Press, 1987), chap. 1. R. Descartes, Meditations on First Philosophy (1641), trans. and ed. J. Cottingham

(Cambridge: Cambridge University Press, 1996), p. 21. 3 J.L. Mackie, The Cement of the Universe (Oxford: Oxford University Press, 1974), pp. 12–13. 4 In fact, Barry Stroud’s interpretation (which I class as a version of the traditional interpretation) does not deny (2); in Stroud’s view, we simply do, inevitably, ‘come to believe, mistakenly, that there are objective necessary connections between events’ (B. Stroud, Hume (London: Routledge, 1977), p. 87). So Stroud apparently embraces, rather than attempts to resolve, the tension described above. 5 Stroud, Hume, p. 86. 6 H. Noonan, Hume on Knowledge (London: Routledge, 1999), p. 142. 7 For more on this, see my Hume on Causation (Abingdon: Routledge, 2006), sect. 4.3. 8 See, for example, J.A. Robinson, ‘Hume’s Two Definitions of “Cause”’, Philosophical Quarterly 12 (1962), pp. 162–71; A. Coventry, Hume: A Guide for the Perplexed (London: Continuum, 2007), p. 110; G. Dicker, Hume’s Epistemology and Metaphysics (London: Routledge, 1998), p. 115; Noonan, Hume on Knowledge, p. 151. 9 D. Garrett, Cognition and Commitment in Hume’s Philosophy (New York: Oxford University Press, 1997), chap. 5. 10 Garrett, Cognition and Commitment, p. 108. 11 Craig, The Mind of God, p. 108; see also Garrett, Cognition and Commitment, p. 106. 12 For more on this, see my Hume on Causation, pp. 94–107. 13 Interpretations that come under this general heading include T.L. Beauchamp and A. Rosenberg, Hume and the Problem of Causation (New York: Oxford University Press, 1981); Stroud, Hume; Mackie, The Cement of the Universe; Garrett, Cognition and Commitment. There are important differences between the interpretations offered by these authors, however, on the troublesome issue of what to do about the idea of necessary connection; see my Hume on Causation, chap. 5, for discussion. 14 The major contributions to the sceptical realist literature have been N. Kemp Smith, The Philosophy of David Hume (London: Macmillan, 1941); J.P. Wright, The Sceptical

144

9780826443595_Ch05_Final_txt_print.indd 144

2/9/2012 9:35:03 PM

CAUSATION AND NECESSARY CONNECTION

15

16

17 18 19

20

21

22 23 24

Realism of David Hume (Manchester: Manchester University Press, 1983); G. Strawson, The Secret Connexion: Causation, Realism, and David Hume (New York: Oxford University Press, 1989); S. Buckle, Hume’s Enlightenment Tract (Oxford: Oxford University Press, 2001); and P.J.E. Kail, Projection and Realism in Hume’s Philosophy (Oxford: Oxford University Press, 2007). Wright, The Sceptical Realism of David Hume, p. 147. Kail takes the same line, holding that the ‘causal necessity of which we may be ignorant can be understood as that which, were we to be acquainted with it, would yield a priori inference and render it impossible to conceive cause without effect’ (Kail, Projection and Realism, p. 103). Quoted in Wright, The Sceptical Realism of David Hume, p. 140. Strawson, The Secret Connexion, p. 157. Ibid., p. 126. See A.J. Jacobson, ‘From Cognitive Science to a Post-Cartesian Text: What Did Hume Really Say?’, in R. Read and K. Richman (eds), The New Hume Debate (London: Routledge, 2000), pp. 156–66. See K. Winkler, ‘The New Hume’, in R. Read and K. Richman (eds), The New Hume Debate, pp. 52–87, and S. Blackburn, ‘Hume and Thick Connexions’, in R. Read and K. Richman (eds), The New Hume Debate, pp. 100–12. See my Hume on Causation, pp. 180–93 for further discussion of both options, and Kail, Projection and Realism, pp. 118–21, for more critical assessment. Wright, The Sceptical Realism of David Hume, p. 95. Strawson, The Secret Connexion, pp. 132–3. Ibid., p. 87. The projectivist interpretation is suggested by some remarks of Simon Blackburn’s; see, for example, S. Blackburn, Spreading the Word (Oxford: Oxford University Press, 1984), pp. 210–12 and ‘Hume and Thick Connexions’, pp. 107–11. It is given a fuller treatment in my Hume on Causation, chap. 6, and in A. Coventry, Hume’s Theory of Causation: A Quasi-Realist Interpretation (London:

25 26

27 28 29

30

31

32 33

34

35

36

37 38

39

Continuum, 2006). For a general discussion of the notion of ‘projection’ see Kail, Projection and Realism, chap. 1, and for a discussion of Hume’s (allegedly problematic) attitude towards ‘projection’, see B. Stroud, ‘“Gilding or Staining” the World with “Sentiments” and “Phantasms”’, in R. Read and K. Richman (eds), The New Hume Debate, pp. 16–30. Coventry, Hume’s Theory of Causation. M. Smith, ‘Objectivity and Moral Realism: On the Significance of the Phenomenology of Moral Experience’, in J. Haldane and C. Wright (eds), Reality, Representation, and Projection (New York: Oxford University Press, 1993), p. 246. D. Hume, ‘Of the Standard of Taste’, E 230. Ibid. See Coventry, Hume’s Theory of Causation, pp. 117–33, for further discussion. See, however, P.J.E. Kail, ‘Book Review: Helen Beebee, Hume on Causation’, Mind 117 (2008), pp. 453–4, for a worry about the use of the notion of projection here. See Coventry, Hume’s Theory of Causation, pp. 133–7. See my Hume on Causation, pp. 160–7. The same is generally true of the moral and aesthetic cases, although, as Coventry notes, in ‘Of the Standard of Taste’, Hume twice appears to count matters for which there is a ‘standard’ (as in ‘standard of taste’) as ‘matters of fact’; see Coventry, Hume’s Theory of Causation, p. 119, and Hume, E 230, 242. See my Hume on Causation, pp. 152–4, for more discussion in the context of Hume’s distinction between reason and taste. P. Van Inwagen, An Essay on Free Will (Oxford: Clarendon Press, 1983), p. 65. J. Earman, A Primer on Determinism (Dordrecht: D. Reidel, 1986), p. 5. See my Hume on Causation, sect. 6.5. For a defence of the latter strategy, see G. Strawson, ‘David Hume: Objects and Power’, in R. Read and K. Richman (eds), The New Hume Debate, pp. 31–51. For brief discussion, see my Hume on Causation, pp. 221–5. Wright, The Sceptical Realism of David Hume, p. 147.

145

9780826443595_Ch05_Final_txt_print.indd 145

2/9/2012 9:35:03 PM

6 HUME ON SCEPTICISM AND THE MORAL SCIENCES Alan Bailey

1. HUME’S RADICAL SCEPTICISM The title of the concluding part of Book One of Hume’s Treatise is ‘Of the sceptical and other systems of philosophy’, and Hume wastes no time in setting before the reader two sceptical arguments that attack, when taken in tandem, the justified status of the overwhelming majority of our beliefs. In the opening section, ‘Of scepticism with regard to reason’, the argument outlined by Hume is presented as targeting all beliefs that are the product of any kind of inference. According to this argument, both demonstrative and probabilistic inferences are ultimately unsuccessful in giving their conclusions any positive degree of justification (see THN 1.4.1.1–6 / 180–3). The attempt to found beliefs about mindindependent physical objects on the senses rather than some form of inference then comes under attack in the very next section of the Treatise. In ‘Of scepticism with regard to the senses’, Hume rapidly dismisses the suggestion that the senses are capable by themselves of revealing that some of the objects seemingly encountered by us in perception continue to exist even after they have ceased to appear to the senses: any such supposition is ‘a contradiction in terms, and

supposes that the senses continue to operate, even after they have ceas’d all manner of operation’ (THN 1.4.2.3 / 188). And he follows this with a repudiation of the supposition that the senses are capable of persuading us that some of the things of which we are immediately aware in perception have the ability to exist independently of perception or are representations of entities of another kind that do possess such a power (see THN 1.4.2.4–14 / 189–93). It seems, therefore, that if we take Hume as endorsing the arguments that he has set out in such detail in these key sections of the Treatise, then we must interpret him as accepting that no claim about mind-independent objects and no claim that requires support from any form of inference is ever any better justified than the logical contradictory of that claim. And as Hume appears to hold that the inner and outer senses are unable, without some form of inferential supplementation, to tell us anything about events that are not happening to us at the present moment, this implies that Hume holds that only beliefs about very simple necessary truths that do not need to be grasped through a process of inference and beliefs about the content of our present ideas

146

9780826443595_Ch06_Final_txt_print.indd 146

2/9/2012 9:35:12 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES and impressions are potentially capable of being rationally justified beliefs. It will be convenient from this point onwards to refer to these few beliefs that Hume is seemingly prepared to accept as justifiable as H-minimal beliefs. Now it seems evident that an inquirer who has engaged in substantial epistemological reflection but nevertheless fails to accept that any beliefs other than H-minimal beliefs are rationally justified beliefs is someone who has embraced a radically sceptical posture. It appears, therefore, that we can legitimately reject the conclusion that Hume is himself a radical sceptic only if we can uncover substantial grounds for supposing that Hume does not genuinely endorse the arguments that he has chosen to elaborate in ‘Of scepticism with regard to reason’ and ‘Of scepticism with regard to the senses’. There must be a presumption, however, that an author who puts lengthy arguments before the reader without explicitly distancing himself from those arguments does endorse their conclusions. And it is difficult to see Hume’s comments in the paragraph that brings these two sections to an end as anything other than an attempt to make it clear that he stands firmly behind his negative epistemological arguments:

positive view of sceptical argumentation. Immediately after Hume has set out the argument that no process of inference succeeds in conferring justification on its conclusions, he specifically addresses the question of whether he sincerely gives his assent to this argument and whether he is ‘really one of those sceptics, who hold that all is uncertain, and that our judgement is not in any thing possest of any measures of truth and falshood’ (THN 1.4.1.7 / 183). And his answer is one that could easily be interpreted as a repudiation of the argument that he has just constructed: I shou’d reply, that this question is entirely superfluous, and that neither I, nor any other person was ever sincerely and constantly of that opinion. Nature, by an absolute and uncontroulable necessity has determin’d us to judge as well as to breathe and feel; nor can we any more forbear viewing certain objects in a stronger and fuller light, upon account of their customary connexion with a present impression, than we can hinder ourselves from thinking as long as we are awake, or seeing the surrounding bodies, when we turn our eyes towards them in broad sun-shine. (ibid.) Commentators who doubt the propriety of classifying Hume as a radical sceptic also tend to seize on his explanation of his purposes in presenting his master argument against inferential justification. He indicates that this argument has been included in the Treatise as a way of adding credibility to his preferred account of belief as something that arises when ideas acquire additional force and vivacity as a result of their associative links with impressions:

’Tis impossible upon any system to defend either our understanding or our senses; and we but expose them farther when we endeavour to justify them in that manner. As the sceptical doubt arises naturally from a profound and intense reflection on those subjects, it always encreases, the farther we carry our reflections, whether in opposition or conformity to it. (THN 1.4.2.57 / 218) It does have to be conceded that there are other passages in the Treatise that initially seem to indicate that Hume has a much less

My intention then in displaying so carefully the arguments of that fantastic sect

147

9780826443595_Ch06_Final_txt_print.indd 147

2/9/2012 9:35:12 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES [the sceptics], is only to make the reader sensible of the truth of my hypothesis, that all our reasonings concerning causes and effects are deriv’d from nothing but custom; and that belief is more properly an act of the sensitive, than of the cogitative part of our natures. (THN 1.4.1.8 / 183)

In similar fashion, Hume’s discussion of our belief in the existence of mind-independent physical objects sees him surrounding his presentation of the relevant sceptical arguments with observations that generate significant perplexity about his personal attitude towards such argumentation. Thus we find that the paragraph that opens ‘Of scepticism with regard to the senses’ includes the following declamation: We may well ask, What causes induce us to believe in the existence of body? but ’tis in vain to ask, Whether there be body or not? That is a point, which we must take for granted in all our reasonings. (THN 1.4.2.1 / 187) It might be suggested, accordingly, that if we must indeed hold as a presupposition of all our reasoning that mind-independent bodies exist, this precludes anyone from sincerely endorsing sceptical arguments on this topic. If we are genuinely committed to the conclusion that mind-independent bodies exist, how can we simultaneously attach any weight to arguments that purport to call that conclusion into question? It is apparent, therefore, that the manner in which Hume expounds his philosophical position in the Treatise does place some difficulties in the way of the conclusion that this work incorporates sceptical arguments that have genuinely persuaded him that only

H-minimal beliefs ever possess any positive degree of rational justification. But when we look more closely at his explicit reservations about sceptical argumentation, we discover that he seems intent on distancing himself only from the supposition that such argumentation is capable of radically reshaping our non-epistemic beliefs. He is keen to emphasize that no amount of reflection on sceptical arguments can permanently dislodge everyday beliefs about the existence and properties of such things as trees, chairs and books. This does not mean, however, that he holds that sceptical arguments permit us to regard such beliefs as rationally justified. As we have seen, Hume is sometimes prepared to insinuate that it would be appropriate to regard him as sincerely assenting to the argument that no beliefs based on inference are ever rationally justified only if that argument were to cause him to abandon all inferential beliefs. Thus we find that Hume presents our supposed inability to ‘forbear viewing certain objects in a stronger and fuller light, upon account of their customary connexion with a present impression’ as sufficient of itself to guarantee that it is impossible for anyone to give their sincere assent to this argument (THN 1.4.1.7 / 183). It would normally be supposed, though, that a distinction can be drawn between sincere assent to the conclusion that a particular belief is not a rationally justified belief and the actual abandonment of that belief. Many people suffering from acrophobia, for example, sincerely accept that at least some of their fears about being in high places are not rationally justified fears. Nevertheless this does not prevent them from continuing to be afflicted by those fears. It seems, accordingly, that Hume’s insistence that neither he nor anyone else is capable of eschewing all inferential beliefs in response to the argument set

148

9780826443595_Ch06_Final_txt_print.indd 148

2/9/2012 9:35:12 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES out in ‘Of scepticism with regard to reason’ is entirely compatible with the supposition that he sincerely assents to that argument in the sense that he endorses as true its conclusions about the limitations on our ability to arrive at epistemically justified beliefs. This latter interpretation of Hume’s position gains additional credibility from other remarks that he makes about the argument against reason: these remarks are very difficult to explain if we are not prepared to regard him as holding that only someone who has already embraced the view that scarcely any of our beliefs are rationally justified is in a position to deny, in good faith, that this argument provides rationally compelling support for just that assessment of our epistemic situation. Hume boldly claims, for instance, that the argument against reason proves that the principles that lead us to correct our judgements on any subject through ‘the consideration of our genius and capacity, and of the situation of our mind, when we examin’d that subject’ would, if we persisted in following them, reduce all putative evidence to nothing and ‘utterly subvert all belief and opinion’ (THN 1.4.1.8 / 184). He also explicitly bases his affirmation that we can safely conclude that ‘reasoning and belief is some sensation or peculiar manner of conception, which ’tis impossible for mere ideas and reflections to destroy’ (ibid.) on his confidence that no one will be able to find any error in the arguments contained in ‘Of scepticism with regard to reason’. It seems clear, therefore, that when we consider the foregoing pieces of corroborative evidence in combination with Hume’s comments in the final section of Book One of the Treatise that he has shown, by means of the argument against reason, that the understanding entirely undermines itself when it operates alone and that we are left with no choice ‘but

betwixt a false reason and none at all’ (THN 1.4.7.7 / 268), we have compelling grounds for asserting that this argument’s conclusion about the extremely limited availability of rationally justified belief does indeed enjoy Hume’s full support. This same pattern is also discernible within Hume’s discussion of the status of our beliefs about the existence of mind-independent physical objects. It is indeed true that he emphatically repudiates the supposition that any of his arguments could bring about more than a fleeting alienation from those beliefs (see THN 1.4.2.57 / 218). Nevertheless it seems plain that he regards the sceptical attack on their supposed rational credentials as completely unanswerable. Instead of diagnosing this sceptical critique as relying on some fallacious inferential step or a premise that is either false or readily deniable, Hume affirms that sceptical worries about the justification of the supposition that mind-independent physical objects exist are irrefutable: ‘This sceptical doubt, both with regard to reason and the senses, is a malady, which can never be radically cur’d, but must return upon us every moment, however we may chace it away, and sometimes may seem entirely free from it’ (ibid.). From Hume’s perspective, no process of reflective thought can place that supposition on a justified basis, and the intense examination of sceptical arguments, even if undertaken with the intention of refuting them, simply makes their irrefutability more salient and potentially strengthens their psychological impact to such an extent that they can fleetingly alienate us from our common-sense belief in physical objects. Hume accordingly rejects the project of constructing a philosophical answer to scepticism in favour of delineating a way of arriving at a psychological accommodation with scepticism that

149

9780826443595_Ch06_Final_txt_print.indd 149

2/9/2012 9:35:13 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES serves to minimize its potentially disruptive effects. As philosophical inquiry cannot locate any deficiencies in the arguments of the sceptics that are not shared by all other forms of reasoning, Hume suggests that the best way of responding to these epistemological difficulties is to relax our intellectual efforts: ‘Carelessness and in-attention alone can afford us any remedy’ (ibid.). It seems, therefore, that the most plausible interpretation of the explicit discussions of sceptical argumentation in the Treatise is that Hume combines the conviction that these arguments make it impossible for him, or indeed any inquirer who has fully understood their implications, to maintain in good faith that beliefs other than H-minimal beliefs are ever rationally justified with the judgement that arriving at this conclusion has little psychological impact on our everyday, non-epistemic beliefs. However, the credibility of this assessment of Hume’s stance also depends on the extent to which it fits with his presentation of his views in the Enquiry concerning Human Understanding. And we find, in fact, that this work displays numerous elements that prevent us from regarding it as less sceptical than Book One of the Treatise. Several arguments deployed in the Treatise are explicitly acknowledged by Hume as sceptical arguments only when they appear in the Enquiry concerning Human Understanding. Consider, for example, the revisions he makes to the presentation of his analysis of causal reasoning. Not only does he give the initial stages of this analysis the forthright title ‘Sceptical Doubts concerning the Operations of the Understanding’, but we also find that the immediately following section begins with a discussion of ‘the academical or sceptical philosophy’ in which he says that such a philosophy is harmless and innocent

in almost every instance and attempts to reassure us that our natural instincts are sufficiently powerful to ensure that it can never ‘undermine the reasonings of common life, and carry its doubts so far as to destroy all action, as well as speculation’ (EHU 5.1–2 / 41). Moreover, the piece of abstract reasoning Hume displays as an example of the kind of argument held in check by these natural instincts is nothing other than a summary of the argument offered in Section 4 (See EHU 5.2 / 41). And even more striking is the manner in which Hume shapes the discussion of scepticism in the final section of the Enquiry. In Part 2 of Section 12 he uses the terms ‘Pyrrhonism’ and ‘excessive scepticism’ to refer to a form of scepticism that attempts to ‘destroy reason by argument and ratiocination’ (EHU 12.17 / 155).1 According to him, this form of scepticism purports ‘to find objections, both to our abstract reasonings, and to those which regard matter of fact and existence’ (EHU 12.17 / 156); and when he comes to describe the most forceful of these sceptical objections to reasonings concerning matters of fact, the argument he ascribes to the Pyrrhonist is essentially a reprise of Hume’s own negative claims about the grounding of causal inferences. In addition to the foregoing greater willingness to identify his own arguments as sceptical arguments, further evidence within the Enquiry of Hume’s commitment to some form of radical scepticism is furnished by his discussion of the relationship between Pyrrhonean and Academic scepticism. In the final section of the Enquiry Hume seems to imply that the only significant criticism that can be levelled against the arguments used by Pyrrhonean sceptics is that they lack the power to dislodge our non-epistemic beliefs (see EHU 12.23 / 159–60). And he indicates that the causal resultant of natural instinct

150

9780826443595_Ch06_Final_txt_print.indd 150

2/9/2012 9:35:13 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES and Pyrrhonean scruples about rational justification is an enduring and advantageous disposition to form one’s non-epistemic beliefs in a cautious and undogmatic manner. According to Hume, the most convenient way of acquiring this disposition is ‘to be once thoroughly convinced of the force of the pyrrhonian doubt, and of the impossibility, that any thing, but the strong power of natural instinct, could free us from it’ (EHU 12.25 / 162). Thus mitigated scepticism or the Academical philosophy, which is strongly recommended to the reader in both Sections 5 and 12 of the Enquiry, is presented not as the outcome of recognizing flaws in the arguments of radical sceptics but rather as the outcome of our natural belief-forming mechanisms causally interacting with the realization that sceptical arguments can only be repudiated in good faith by someone who already accepts that no beliefs other than H-minimal beliefs are ever rationally justified.2 Ultimately, then, the considerations in favour of interpreting Hume as committed to a radically sceptical stance in both the Treatise and the Enquiry concerning Human Understanding seem very strong indeed. Prominent arguments in Book One of the Treatise terminate in the conclusion that no beliefs other than H-minimal beliefs are rationally justified; and once we have given due weight to the distinction between seeing a belief as devoid of rational justification and actually abandoning that belief, there seems to be little explicit basis in the text for supposing that Hume does not endorse that conclusion. When, therefore, we find a similarly radical array of sceptical arguments deployed in the Enquiry alongside seemingly robust affirmations of their strength and irrefutability (see EHU 12.14 / 153–4, 12.22 / 159, and 12.25 / 162), it surely ceases to be plausible

to hold that Hume does not embrace the sceptical view that very few of our beliefs are ever rationally justified.

2. THE TENSION BETWEEN SCEPTICISM AND ABSTRUSE INQUIRY As we have just seen, discounting Hume’s commitment to a radically sceptical perspective in respect of the limited availability of rationally justified belief does not appear to be a viable interpretative strategy. However, if we accept that reflection on sceptical arguments has led Hume to the conclusion that only H-minimal beliefs can ever qualify as rationally justified beliefs, there does seem to be a real danger that Hume’s overall intellectual posture ceases to be a coherent one. The subtitle of the Treatise is Being an Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects. And in the introduction to the Treatise, Hume presents this work as primarily an attempt to explain the best way of making progress in the science of man. Hume maintains that the science and the philosophy of his era are replete with systems and theories that display such undesirable qualities as ‘principles taken upon trust, consequences lamely deduc’d from them, want of coherence in the parts, and of evidence in the whole’ (THN Intro. 1 / xiii). These glaring deficiencies ‘seem to have drawn disgrace upon philosophy itself’ (ibid.), and they have also given rise to what Hume describes as ‘that common prejudice against metaphysical reasonings of all kinds, even amongst those, who profess themselves scholars, and have a just value for every other part of literature’ (THN Intro. 2 / xiv). As he explains, this prejudice against metaphysical reasonings is

151

9780826443595_Ch06_Final_txt_print.indd 151

2/9/2012 9:35:13 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES not simply a prejudice against a particular branch of inquiry but has come to constitute a general aversion to ‘every kind of argument, which is in any way abstruse, and requires some attention to be comprehended’ (ibid.). Hume’s declared ambition is to alleviate this prejudice by helping to put science and philosophy on a sound basis. His proposed way of doing this involves applying the methods of the natural sciences, as exemplified within the most compelling theories that have arisen within that realm of inquiry, to the study of human nature. Hume holds that once we have a better understanding of the mental activities of human beings, this will shed fresh light on a wide variety of intellectual disciplines: ’Tis evident, that all the sciences have a relation, greater or less, to human nature; and that however wide any of them may seem to run from it, they still return back by one passage or another. Even Mathematics, Natural Philosophy,3 and Natural Religion, are in some measure dependent on the science of man; since they lie under the cognizance of men, and are judg’d of by their powers and faculties. (THN Intro. 4 / xv) Of course natural philosophy is the arena in which Hume has found examples of convincing theories built on a credible investigative methodology, an investigative methodology he has deliberately extended to the moral sciences. His strategy, therefore, involves making use of a virtuous cycle of mutual improvement. Such exemplary theories within the natural sciences as ‘newton’s explication of the wonderful phenomenon of the rainbow’ (DNR 1.136) and ‘the copernican system’ (ibid. 138) serve to validate the experimental and experience-based investigative methodology that Hume regards as having given

rise to those theories. This methodology is, in turn, something that can beneficially be applied to the study of human nature and in particular to the study of how human beings make inferences and form judgements. And with the assistance of the ensuing improved account of human psychology, we will be able to develop a still better methodology of inquiry for use not only within the natural sciences themselves but also within other intellectual disciplines that lack acknowledged achievements on a par with the theories of Newton, Galileo and Copernicus. Significantly, Hume is adamant that the progress in the natural sciences that has finally given us some theories that are both intellectually satisfying and stable in the face of examination is the product of the experimental method of investigation. He specifically places his project in the Treatise in the context of an intellectual movement, exemplified in his eyes by ‘some late philosophers in England, who have begun to put the science of man on a new footing, and have engag’d the attention, and excited the curiosity of the public’ (THN Intro. 7 / xvii), that is attempting to transfer the application of experimental philosophy from natural to moral subjects. Hume maintains that this attempt forces us to acknowledge that the science of man can only be built on a foundation ‘laid on experience and observation’ (THN Intro. 7 / xvi). If we construct the science of man in this manner, then our conclusions about human psychology will assist us in refining and improving our investigative techniques across a broad range of subjects. But if we abandon the guidance of experience, then we will be led to conclusions that ‘ought, at first, to be rejected as presumptuous and chimerical’ (THN Intro. 8 / xvii). These forthright scientific and philosophical aspirations, however, raise the question

152

9780826443595_Ch06_Final_txt_print.indd 152

2/9/2012 9:35:13 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES of how they cohere with Hume’s radical scepticism. Given their prominence at the start of the Treatise, how does it come about that Book One seems to terminate in a reaffirmation of sceptical arguments indicating that only beliefs with H-minimal content are ever rationally justified beliefs? No science can be constructed that consists exclusively of beliefs with such attenuated content. Should we conclude, therefore, that Hume’s positive project has catastrophically imploded in the course of writing Book One and that he has unintentionally found himself immersed in a sceptical quagmire from which there is no principled escape? The supposition that Hume sees the outcome of Book One of the Treatise as inconsistent with or even tangential to his initial ambitions to put the study of human nature on a sounder footing is difficult to reconcile with the untroubled opening of Book Two. In Book Two he presents an account of the operations of the human passions that seems perfectly in accord with his professed intentions to make progress in the moral sciences, and there is no indication whatsoever that he regards this account as undermined or in any way rendered problematic by the sceptical reflections developed in the final stages of Book One. Similarly, his discussion of morality in Book Three shows no signs of being infected by destabilizing sceptical worries. It is true that in the opening paragraph he ventures some observations about the limited persuasive force of abstruse reasoning, but the overwhelming tenor of his discussion is one of quiet confidence in the satisfactory nature of his arguments and conclusions: I am not, however, without hopes, that the present system of philosophy will acquire new force as it advances; and that our reasonings concerning morals

will corroborate whatever has been said concerning the understanding and the passions. (THN 3.1.1.1 / 455) We also need to keep in mind that if Hume had seen the sceptical crisis of Book One as an unwelcome or unexpected interruption in the unfolding of his overall line of argument in the Treatise, then he had ample opportunity to omit it from the version of the text sent for publication or to include additional guidance to the reader about how to place this element of his discussion within his positive aims. His letters confirm that he had no qualms about deleting some of the more obviously irreligious material from the text before finally placing it in the hands of his publisher (see, in particular, LDH 1.25, 6). Thus the prominent role ultimately allocated in the Treatise to the presentation of sceptical lines of thought presumably reflects Hume’s considered judgement that the sceptical aspects of his thinking and the attempt to build the science of human nature on a sound foundation form part of one coherent and mutually supportive intellectual enterprise. Once again Hume’s revised presentation of his views in the Enquiry concerning Human Understanding provides important corroboration of what is implied by the text of the Treatise. In Section 1 of the Enquiry Hume develops a defence of ‘profound reasonings, or what is commonly called metaphysics’ and abstract philosophy (EHU 1.7 / 9) that presents it as providing us with, amongst other benefits, an accurate delineation of ‘the powers and faculties of human nature’ (EHU 1.13 / 13). Hume even suggests that the analogy with the development of Newtonian mechanics means that we can legitimately hope that such profound and abstruse reasonings will ultimately serve to ‘discover, at least in some degree, the secret springs and

153

9780826443595_Ch06_Final_txt_print.indd 153

2/9/2012 9:35:13 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES principles, by which the human mind is actuated in its operations’ (EHU 1.15 / 14). We are forced to conclude, therefore, that Hume’s intention to assist in the construction of a science of human nature remained intact at the time of writing the Enquiry despite the effort lavished in the Treatise on the elaboration of sceptical lines of argument. Yet the Enquiry also sees Hume deploying another battery of sceptical arguments that are scarcely any less radical than those found in the Treatise. In some ways, indeed, the Enquiry even seems to place a greater emphasis than the Treatise on the force and tenaciousness of sceptical argumentation. It appears, therefore, that the Enquiry concerning Human Understanding replicates, without any obvious sign of unease or tension on Hume’s part, the same blend of philosophico-scientific ambitions and sceptical conclusions about the limited availability of rationally justified beliefs that can be discerned in the Treatise. Once we accept, however, that Hume regards radical scepticism and the attempt to construct a science of man as complementary activities, it is natural to wonder what kind of science can exist in a context where all the core beliefs of this putative science are viewed by its own adherents as totally devoid of rational justification. Moreover, if the most profound and advanced forms of scientific thinking are as compatible with radical scepticism about justification as Hume seems to think, what does this say about the value of engaging in sceptical reflection? In the final section of the Enquiry concerning Human Understanding, Hume presents what he plainly regards as a potentially devastating criticism of radical scepticism: For here is the chief and most confounding objection to excessive scepticism, that no durable good can ever result from it;

while it remains in its full force and vigour. We need only ask such a sceptic, What his meaning is? And what he proposes by all these curious researches? He is immediately at a loss, and knows not what to answer. (EHU 12.23 / 159–60) Hume explains this objection in terms of the inability of sceptical argumentation to produce stable changes in people’s beliefs and behaviour analogous to those that can arise from exposure to the Copernican system of astronomy or the moral exhortations of the Stoics and Epicureans. According to Hume, the radical or Pyrrhonean sceptic ‘cannot expect, that his philosophy will have any constant influence on the mind: Or if it had, that its influence would be beneficial to society’ (EHU 12.23 / 160). It seems, therefore, that Hume is suggesting that we are entitled to dismiss reflection on sceptical reasoning as pointless or, at best, a frivolous and esoteric form of amusement unless there is some indication that it can alter a person’s set of beliefs in stable and useful ways. Somewhat ironically Hume’s own attempt to emphasize the transient nature of the doxastic changes induced by sceptical arguments actually terminates by drawing attention to one stable change to which these arguments can give rise. He insists that despite the initial impact of the sceptic’s arguments, it is evident that ‘the first and most trivial event in life will put to flight all his doubts and scruples’ (ibid.). Thus he describes the sceptic as awakening from a dream and laughing at himself for being swayed by these abstruse epistemological objections. Nevertheless this reimmersion in quotidian life does not entirely negate the influence of sceptical reasoning. The sceptic may indeed be led to describe his arguments as nothing more than amusing puzzles, but Hume also represents the sceptic

154

9780826443595_Ch06_Final_txt_print.indd 154

2/9/2012 9:35:14 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES as retaining a distinctive perspective on the rationality of his and our beliefs. Even when the sceptical fugue that can be generated by prolonged reflection on epistemological arguments passes, the sceptic retains a robust sense of an aspect of the human predicament that escapes the attention of more superficial inquirers: the whimsical condition of mankind, who must act and reason and believe; though they are not able, by their most diligent enquiry, to satisfy themselves concerning the foundation of these operations, or to remove the objections, which may be raised against them. (ibid.) Given that Hume draws our attention to this enduring consequence of the careful study of sceptical reasonings, why does he also make the seemingly contradictory claim that radical scepticism fails to exercise any constant influence on the mind? The most plausible answer here is that he is thinking in terms of influence over our everyday nonepistemic beliefs. The radical sceptic might well acquire a settled disposition to deny that any beliefs other than H-minimal beliefs are ever rationally justified beliefs, but this does not lead to suspension of belief on the existence of cattle, chairs and chimneys. Indeed the radical sceptic’s expectations about what will happen, unlike his beliefs about what we are justified in thinking will happen, tend to be remarkably similar to our own. Thus the objection to radical scepticism that Hume emphasizes in the Enquiry concerning Human Understanding seems to amount to the contention that if the argumentative case assembled so laboriously by sceptical thinkers has no or almost no impact on our non-epistemic beliefs, then we have no motive for expending the effort required to

follow such reasoning. Some rather unusual people might find that the study of sceptical argumentation serves them as a mildly entertaining pastime, but Cleanthes undoubtedly speaks for most of us when he expresses a preference in the Dialogues concerning Natural Religion for less demanding and cerebral forms of amusement: ‘But for my part, whenever I find myself disposed to mirth and amusement, I shall certainly choose my entertainment of a less perplexing and abstruse nature. A comedy, a novel, or at most a history, seems a more natural recreation than such metaphysical subtilties and abstractions.’ (DNR 1.137) It might be suggested at this point that the sceptic can mount an effective response to Hume’s objection by drawing a distinction between the non-epistemic beliefs that are an integral part of everyday life and the nonepistemic beliefs generated by such activities as philosophy and theoretical science. Even if these everyday beliefs cannot be disturbed or changed by sceptical argumentation, perhaps the arguments deployed by radical sceptics can reshape our philosophical and scientific commitments. Suspension of judgement at the everyday level is certainly not something that flows from reflection on sceptical reasonings, but that does not guarantee that the more abstract and sophisticated nonepistemic beliefs found within philosophy and the sciences are similarly invulnerable to sceptical attack. This indeed is Philo’s initial response to Cleanthes’ charge that his professed scepticism is an insincere affectation. Cleanthes challenges Philo to exhibit his scepticism in his behaviour: Whether your scepticism be as absolute and sincere as you pretend, we shall learn bye and bye, when the company

155

9780826443595_Ch06_Final_txt_print.indd 155

2/9/2012 9:35:14 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES breaks up: We shall then see, whether you go out at the door or the window; and whether you really doubt, if your body has gravity, or can be injured by its fall; according to popular opinion, derived from our fallacious senses and more fallacious experience. (DNR 1.132)

In this crude form Cleanthes’ attack is easily parried by Philo. Philo simply appeals to the involuntary nature of these everyday beliefs. The sceptic sincerely and genuinely holds that even straightforward and seemingly uncontroversial everyday beliefs are entirely devoid of rational justification. However, this conviction fails to dislodge these beliefs because they are held in place by psychological forces that are causally impervious to any influence that can be exerted by sceptical arguments: ‘To whatever length any one may push his speculative principles of scepticism, he must act, I own, and live, and converse like other men; and for this conduct he is not obliged to give any other reason than the absolute necessity he lies under of so doing.’ (DNR 1.134) Philo also wishes, however, to leave room for radical scepticism to affect at least some of our non-epistemic beliefs in a potentially beneficial way. Consequently he suggests that abstruse theoretical beliefs can be subverted by sceptical arguments: But it is evident, whenever our arguments . . . run wide of common life, that the most refined scepticism comes to be upon a footing with them, and is able to oppose and counterbalance them. The one has no more weight than the other. The mind must remain in suspense between them; and it is that very suspense or balance, which is the triumph of scepticism. (DNR 1.135–6)

This second component of Philo’s defence of scepticism could be construed as overlooking the possibility that refined arguments that run wide of common life might sometimes be arguments that should be allowed to influence us. However, it would be more charitable to interpret Philo as attempting to draw a distinction between complicated and intricate arguments that are nevertheless rooted in the same underlying principles as those that generate stable conviction in ordinary life and even more ambitious and speculative arguments that disregard those common-sense principles. Thus we find that Philo is happy to endorse a form of theorizing in both the natural and moral sciences that keeps closely to patterns of inference and evidence assessment that exert a major influence on our quotidian beliefs: He considers besides, that every one, even in common life, is constrained to have more or less of this philosophy; that from our earliest infancy we make continual advances in forming more general principles of conduct and reasoning; that the larger experience we acquire, and the stronger reason we are endowed with, we always render our principles the more general and comprehensive; and that what we call philosophy is nothing but a more regular and methodical operation of the same kind. (DNR 1.134) It seems, therefore, that Philo’s response to the question of the practical value of reflection on sceptical arguments is to claim that such reflection is useful because it serves to subvert and sweep away speculative beliefs that are not rooted in the core psychological mechanisms that generate our enduring everyday beliefs. At the same time this reflection is not dangerous because those everyday beliefs and even the more sophisticated beliefs that

156

9780826443595_Ch06_Final_txt_print.indd 156

2/9/2012 9:35:14 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES arise from the methodical and systematic application of common-sense principles of inference are founded in permanent features of human nature that cannot be overturned by any form of abstract reasoning. Is it really the case, however, that we can satisfactorily combine the view that radical scepticism is intellectually safe with the supposition that it is genuinely useful? In the Dialogues Cleanthes returns to the attack by alleging that self-professed sceptics invariably seem to conduct even their theoretical inquiries in the same manner as people who do not regard themselves as sceptics: But I observe . . . with regard to you, philo, and all speculative sceptics, that your doctrine and practice are as much at variance in the most abstruse points of theory as in the conduct of common life. Wherever evidence discovers itself, you adhere to it, notwithstanding your pretended scepticism; and I can observe, too, some of your sect to be as decisive as those who make greater professions of certainty and assurance. (DNR 1.136)

appears to leave him potentially vulnerable to the charge that these views turn out to have no practical consequences whatsoever. If the radical sceptic responds at the level of his non-epistemic beliefs to what he would call ‘putative’ or ‘alleged’ evidence in exactly the same manner that the non-sceptic responds to what he views as real evidence, then it seems that even if we disregard the issue of sincerity, there is nothing of any substantial value to be gained from the mental exertion required to sustain a sceptical stance. On the other hand, if the radical sceptic genuinely finds, as Philo suggests, that his mental posture is modified so that refined and abstruse reasoning loses much of its usual persuasiveness, how can a radical sceptic be convinced by the refined and abstruse reasoning that underpins Hume’s science of human nature?

3. SCEPTICISM AND THE CURBING OF THE RESTLESS IMAGINATION

It is now possible to appreciate the full force of Cleanthes’ attack on radical scepticism. Initially it appeared that he was concentrating on the issue of whether a person can sincerely embrace the thesis that no beliefs other than H-minimal beliefs are ever rationally justified. And Philo’s appeals to psychological necessitation seemed to provide a satisfactory explanation of why the sceptic shares so many non-epistemic beliefs with people who would vehemently repudiate radical scepticism. However, Philo’s ensuing attempt to locate an area of inquiry in which the radical sceptic’s distinctive views about the limited availability of rationally justified beliefs generate a plausibly beneficial reordering of his non-epistemic beliefs

It is clear that a satisfactory explanation of how Hume sees scepticism and the science of human nature as mutually reinforcing components of a genuinely illuminating worldview needs to do more than merely point to his commitment to the supposition that many of our beliefs are the inevitable causal products of psychological mechanisms that cannot be disrupted by abstract reasoning. That commitment accounts for Philo’s insouciance in the face of Cleanthes’ misguided contention that a true sceptic would have no motive for leaving via the door and stairs rather than a high window. But it does not explain how radical scepticism can give rise to real changes amongst a person’s non-epistemic beliefs while leaving refined and abstruse reasoning with enough persuasiveness to support

157

9780826443595_Ch06_Final_txt_print.indd 157

2/9/2012 9:35:14 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES an edifice of theory with the complexity manifested by the science of man that Hume is attempting to develop in the Treatise. In order to find Hume’s explicit account of how this is possible, we must turn our attention to the final section of Book One. Hume appears content in this section to present sceptical arguments as serving to expose our lack of rational justification for our beliefs. Indeed, he even admits that these arguments can fleetingly alienate us from our most stable beliefs about contingent matters of objective fact. He also emphasizes that ‘the intense view’ of the ‘manifold contradictions and imperfections in human reason’ has effects that cannot be dispelled through the efforts of reason itself (THN 1.4.7.8 / 268–9). Nature, rather than reason, ultimately brings such doxastic alienation to an end, but the initial outcome of this fortuitous cure is a state of ‘spleen and indolence’ (THN 1.4.7.11 / 270) that is not conducive to any further philosophical or scientific inquiry. As time passes, however, and the inquirer grows ‘tir’d with amusement and company’ (THN 1.4.7.12 / 270), an inclination to return to intellectually taxing investigations spontaneously arises again, and new attempts at constructing intellectual systems begin. Hume is adamant, however, that this process operates in accordance with a natural cycle. Until such time as a psychological propensity to find some principles and claims immediately appealing and convincing is activated, and the mind also recovers its enthusiasm for intricate reasoning falling outside the boundaries of common life, it is simply not psychologically possible to arrive at any fresh beliefs in either the moral or the natural sciences in the face of the influence exerted by remembered sceptical arguments. But once these mechanisms are in play once more, the construction of a science of man in accordance

with the psychological foundations provided by experience and observation can resume: If the reader finds himself in the same easy disposition, let him follow me in my future speculations. If not, let him follow his inclination, and wait the returns of application and good humour. (THN 1.4.7.14 / 273) Hume describes the conduct of a man who engages in philosophical inquiry when the inclination spontaneously arises within him, or in ‘this careless manner’, as ‘more truly sceptical than that of one, who feeling in himself an inclination to it, is yet so over-whelm’d with doubts and scruples, as totally to reject it’ (ibid.).4 The point here is that resisting one’s natural propensities is something that requires a rationale. Sceptical arguments, however, subvert standard epistemic reasons: it is no longer possible for a sceptic, even of a mitigated or Academic variety, to think of himself as striving against his inclinations in order to construct a set of beliefs in accordance with the demands of reason. Someone who strenuously strives to shed his beliefs on the grounds that this is what reason demands of an inquirer who no longer views his beliefs as rationally justified has failed to appreciate that this alleged constraint of reason is itself something that sceptical arguments expose as devoid of rational justification. Now a suspensive sceptic might possibly claim that he finds himself convinced that only widespread suspension of belief can protect him against erroneous beliefs that are not balanced by equally valuable true beliefs, and that this constitutes a motive for suspending judgement. Hume’s reply to this contention would presumably be that the most plausible psychological theories available to us imply that sceptical arguments, as opposed perhaps

158

9780826443595_Ch06_Final_txt_print.indd 158

2/9/2012 9:35:14 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES to insanity, cannot subvert our belief-forming mechanisms so completely. Moreover, he can also point out that there is no recorded instance of anyone arriving at such a state, rather than a state of verbally asserting that he is suspending judgement on a radical scale, as a result of exposure to any kind of philosophical argument, and that the behaviour of this self-professed suspensive sceptic makes it clear that he too has numerous beliefs. Indeed it is difficult to see how anyone, no matter how carried away by intense philosophical reflection, could deny in good faith that he generally believes that if wholesale suspension of belief were to be instantiated, then ‘all human life must perish’ (EHU 12.23 / 160). More plausibly, perhaps, someone who purports to be unwilling to engage in philosophico-scientific theorizing and systembuilding might argue that he simply finds the results of abstruse and complicated reasoning, at least about matters of fact, to be so full of ‘errors and delusions’ (THN Intro. 3 / xiv) that he sees no point in continuing with an enterprise that has such a disappointing outcome. Even this position, however, seems an untenable one. Cleanthes’ comments in the Dialogues are particularly apposite here: ‘In reality, would not a man be ridiculous, who pretended to reject newton’s explication of the wonderful phenomenon of the rainbow, because that explication gives a minute anatomy of the rays of light; a subject, forsooth, too refined for human comprehension.’ (DNR 1.136) Very few people are prepared in practice to repudiate all the results of refined and abstruse reasoning, especially where this yields verifiable predictions of novel phenomena or generates new technologies that assist in the manipulation of the world around us. And if such reasoning is perforce embraced in the case of the natural world, what basis

can there be for denying that it might also be successful in uncovering the truth in the moral sciences? The other direction from which pressure falls on a suspensive sceptic who professes to suspend judgement on the results of all abstruse reasoning is his practice of forming numerous opinions within the arena of everyday life. According to Hume, ‘philosophical decisions are nothing but the reflections of common life, methodized and corrected’ (EHU 12.25 / 162). As long as we remain connected in this way to the reasoning we employ and find persuasive outside of our philosophical investigations, there is no sharp division to be drawn between everyday and philosophical reasoning, and hence there is, it might be thought, no principled motive for refusing to participate in philosophical inquiry. After all, if one appeals once more to the existence of unanswerable Pyrrhonean arguments, or indeed Humean genetic arguments for scepticism based on alleged discoveries about the nature of the psychological mechanisms responsible for our beliefs,5 it is an evident fact, and one that cannot be denied in good faith by any philosophical sceptic, that these arguments tell equally against the justification of the everyday beliefs that all sceptics have in profusion and are equally ineffectual at a psychological level against these everyday beliefs and some scientific theories. As Philo contends in his own defence of scepticism: ‘To philosophise on such subjects [natural and moral subjects] is nothing essentially different from reasoning on common life; and we may only expect greater stability, if not greater truth, from our philosophy, on account of its exacter and more scrupulous method of proceeding.’ (DNR 1.134) Now it might be suggested that this argument from the close relationship between

159

9780826443595_Ch06_Final_txt_print.indd 159

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES philosophical reasoning and everyday reasoning overlooks Hume’s own explanation of the inability of radical sceptical arguments to dislodge our everyday non-epistemic beliefs. Hume accepts that these arguments are plausibly seen as continuations of patterns of reasoning that we embrace and respond to in the course of daily life (see, in particular, THN 1.4.1.8–9 / 183–5). However, despite this intimate relationship with common-sense reasoning, they fail to have much impact on our non-epistemic beliefs. Hume explains this in terms of the obstacles the imagination encounters when it attempts to participate in complicated and subtle reasoning: ‘Where the mind reaches not its objects with easiness and facility, the same principles have not the same effect as in a more natural conception of the ideas; nor does the imagination feel a sensation, which holds any proportion with that which arises from its common judgments and opinions’ (THN 1.4.1.10 / 185). An important feature of this explanation, however, is that complication and subtlety are matters of degree. From Hume’s perspective, arguments gradually lose their persuasiveness as their intricacy and complexity increase: this loss of persuasiveness does not stem from a change in the fundamental nature of the argumentation presented to us. Thus the sceptic who maintains that it is appropriate for him to suspend judgement on all conclusions reached by abstruse and refined reasoning might be tempted to dismiss the significance of the close connections between everyday reasoning and the reasoning used in philosophy and science. After all, sceptical reasoning and everyday reasoning also have much in common, but Hume concedes that they affect our non-epistemic beliefs very differently. The problem with this response is that it fails to take into account the fact that some

relatively complicated arguments in the natural sciences do patently command the assent of even the most determined sceptics. As reasoning in the natural and moral sciences is ‘nothing but a more regular and methodical operation of the same kind’ (DNR 1.134) as reasoning in common life, this affords a strong presumption in favour of its being very similar in its persuasiveness to the reasoning that shapes our beliefs in everyday contexts. Now the suspensive sceptic rightly injects a note of caution here. The relationship between everyday reasoning and arguments for radical scepticism conspicuously fails to ensure that sceptical arguments have a similar influence over our non-epistemic beliefs. But when we discover that complicated arguments in the natural sciences do sometimes lead to scientific theories that even the most hardened sceptic would find it ridiculous to reject, then the presumption under discussion here must surely carry a great deal of weight. The only recourse left open to the suspensive sceptic would be to challenge the assimilation of arguments in the human sciences to arguments in the natural sciences rather than radical Pyrrhonean arguments. Arguments in the natural sciences and radical sceptical arguments both manifest a high level of abstraction and complexity. However, some arguments of this kind within the field of the natural sciences generate stable non-epistemic beliefs, yet abstract and complicated arguments of a radically sceptical kind appear to lack this capacity. What, then, explains this difference? Until we understand why the outcomes are so dissimilar, we might be mistaken in treating arguments in the human sciences as potentially performing in the manner of their counterparts in the natural sciences. The explanation that Hume seems to endorse is that arguments in the natural sciences gain persuasiveness and psychological

160

9780826443595_Ch06_Final_txt_print.indd 160

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES force from their strong links with experience and observation: It is only experience, which teaches us the nature and bounds of cause and effect, and enables us to infer the existence of one object from that of another. Such is the foundation of moral reasoning, which forms the greater part of human knowledge, and is the source of all human action and behaviour. (EHU 12.29 / 164) Sceptical arguments, in contrast, are relatively detached from experience, and they represent instead attempts at exploring the implications of various abstract epistemic norms. Hume says little about the origins of these norms, possibly because he believes that all his readers will readily acknowledge both the psychological pull exerted by these norms and the key prescriptions they lay down. For our current purposes, however, the key point is that sceptical arguments seem to rely entirely on such norms’ antecedent psychological force. And as the argumentative links with this original source of vivacity and persuasiveness become increasingly complicated and intricate, the conclusions become less and less convincing. In the case of arguments in the natural sciences, however, their persuasiveness can be reinforced by adding additional lively impressions to our stock of supporting evidence or by securing observations that confirm the predictions of the theories built upon this argumentation or the effectiveness of the technology arising from those theories. The closest that sceptical arguments can come to exploiting such links with lively impressions is when we take ourselves to be encountering impressions that falsify beliefs we have formed on the basis of reasoning that sceptical arguments condemn as

insufficiently rigorous. Repeated experiences of that kind might tend to reinforce our commitment to the standards of self-criticism exploited by radical scepticism. But even this might, in turn, be subverted by a recognition that such self-criticism would also suggest that we are being excessively rash in taking our apparent observations of falsifying phenomena at face value. Essentially, therefore, the Humean view is that compelling scientific arguments and inferences gain persuasiveness from being directly linked to the enlivening effects of sense perception, whereas sceptical arguments have little, if any, scope to exploit this source of conviction. Thus the contest between good scientific arguments and even the most powerful sceptical arguments is, at the non-epistemic level, a distinctly unequal contest: These [sceptical] principles may flourish and triumph in the schools; where it is, indeed, difficult, if not impossible, to refute them. But as soon as they leave the shade, and by the presence of the real objects, which actuate our passions and sensations, are put in opposition to the more powerful principles of our nature, they vanish like smoke, and leave the most determined sceptic in the same condition as other mortals. (EHU 12.21 / 159) When it comes to epistemic appraisals, in contrast, the struggle is one that sceptical arguments can win. It is one thing to be steadfastly convinced that it is true that p and quite another to be convinced that one’s belief that p meets the abstract epistemic standards required for it to qualify as a rationally justified belief. The first level of conviction can be reached simply with the assistance of direct perception or a causal inference based on a lively impression and associative links created by past patterns within one’s impressions.

161

9780826443595_Ch06_Final_txt_print.indd 161

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES However, the second level of conviction requires the belief to be assessed against a set of epistemic norms that are not themselves directly rooted in perception. And as positive assessments are themselves the products of a relatively artificial and strained mental posture, they can be readily counterbalanced by sceptical arguments. The vulnerability of our epistemic appraisals derives from the fact that the realm of such appraisals, as opposed to the realm of beliefs about mere matters of physical and psychological fact, is itself a relatively rarefied arena of thought that has only an indirect connection with what we are told by observation and experience. Because it is not a natural and instinctive turn of thought, conviction is not as strong and as stable as it often is concerning straightforward matters of fact: No wonder, then, the conviction, which arises from a subtile reasoning, diminishes in proportion to the efforts, which the imagination makes to enter into the reasoning, and to conceive it in all its parts. Belief, being a lively conception, can never be entire, where it is not founded on something lively and easy. (THN 1.4.1.11 / 186) It is not just our epistemic beliefs, however, that are vulnerable to sceptical arguments. Non-epistemic judgements that arise from over-lively imaginations that respond to weak associations of ideas, rather than the strong associations that underpin causal inference, are also sufficiently weakly rooted in our natural belief-forming mechanisms that they too can be swept aside by radical sceptical arguments. And it is this vulnerability that allows sceptical argumentation to perform its valuable function of eliminating all those metaphysical positions that are built upon excessively exuberant imaginations and

loose associative links. Theories built upon solid causal reasoning are not similarly vulnerable; so they survive the encounter with even the most powerful sceptical arguments without any substantial damage to their credibility. The only upshot in this latter case is a modest prompting towards a seemly degree of intellectual modesty and a willingness to consult the opinions of other people (see EHU 12.24 / 161–2). Thus the application of radically sceptical arguments, when their power and irrefutability is genuinely internalized, is an extremely efficacious way of orienting the mind towards theories supported by exemplary causal reasoning and away from specious theories supported only by weaker principles of mental association. The outcome for a person who has what Hume regards as the appropriate attitude towards sceptical arguments is that some theory-building continues even after attaining a sceptical apotheosis. Causal arguments that possess substantial experiential backing are almost completely impervious to sceptical arguments at the level of psychological persuasiveness, and they can therefore support conclusions in the natural and human sciences of a recondite and abstruse kind. Engaging in serious intellectual inquiry does, however, leave the inquirer susceptible to sudden shifts in his attitude towards his apparent discoveries. According to Hume, such inquiry will inevitably see us yielding from time to time ‘to that propensity, which inclines us to be positive and certain in particular points, according to the light, in which we survey them in any particular instant’ (THN 1.4.7.15 / 273). Hume sees this as an inevitable psychological consequence of giving our full attention to specific pieces of evidence and reasoning. While no countervailing considerations are before the mind, we are pulled along by the immediate

162

9780826443595_Ch06_Final_txt_print.indd 162

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES plausibility of the narrow and unbalanced range of thoughts present to the mind at that particular moment. No direct effort of will can suppress these momentary lapses into exuberant dogmatism. The only way of eliminating them would be to ‘forbear all examination and enquiry’ (THN 1.4.7.15 / 274). However, the consequences of doing this at the level of everyday enquiry would be, as we noted before, the destruction of all human life (see EHU 12.23 / 160). And although the outcome of eschewing all enquiry of a scientific and philosophical kind would be less disastrous, it would still be absurd for those with an inclination towards such researches to engage in the struggle with their natural impulses that this would involve. The arguments of radical sceptics have the psychological power to force some of us to abandon our initial belief in the existence of certain types of reasons for actions and opinions. However, they do not generate any new reasons to guide people’s behaviour and judgements. They subvert the framework of epistemological rationality, but they put nothing new in its place. Moreover, it would be a serious mistake to suppose that this subversion fails to encompass the supposed normative requirement to eschew beliefs that one cannot sincerely view as rationally justified. Hume’s science of man critiques that alleged norm at a variety of levels. Thus he takes great delight, for example, in repeatedly exposing our psychological inability to put this alleged norm into practice. But the key observation to make in respect of the implications of radical scepticism is that this norm is a self-subverting norm. Suppose that one has been driven, through the contemplation of sceptical arguments, to the conclusion that no beliefs other than H-minimal beliefs are ever rationally justified. It is a commonplace of discussions of

radical scepticism to claim that this is a selfrefuting conclusion. Now it is true that it is, by its own standards, not a rationally justified conclusion. But this does not entail that it is a false conclusion, nor does this give us any grounds for supposing that some contrary conclusion is better justified. Moreover, if it is urged at this point that we are under an intellectual obligation to believe that we ought to eschew beliefs when we cannot regard them as rationally justified, this simply raises the issue of the status of this suspensive norm. And a little reflection seems to establish that it is this norm, at least as it is supposed to hold sway in anyone’s mind, that is, unlike radical scepticism, genuinely self-subverting. Radical scepticism as a set of opinions about the epistemic status of beliefs is fully internally coherent unless it incorporates the suspensive norm currently under discussion, and there is no obvious basis for concluding that it should embrace this norm. Indeed, it appears that there are compelling grounds for supposing that it is not a coherent norm to endorse if one is a radical sceptic. Radical scepticism can be true and we can believe it to be true without transgressing any doxastic norms internal to itself even if it does imply, as a thesis, that it is not a position that possesses any positive justification. But if radical scepticism across the range of topics envisaged by Hume is true and we believe it to be true, then the suspensive norm undermines itself. Let us begin by trying to assume that this norm is true. Then it instructs us to eject it from our minds, as the suspensive norm comes to be its own target given its status as a belief that we cannot regard as justified. On the other hand, if we assume that it is a false norm, then we clearly do not wish to allow it any influence over us. It is, therefore, not a norm we should allow to shape our doxastic

163

9780826443595_Ch06_Final_txt_print.indd 163

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES preferences, because it is either a false norm or a norm that tells us, as radical sceptics, to disregard it. Only a superficial sceptic, accordingly, would combine radical scepticism about all beliefs with more than H-minimal content with the judgement that this places him under some obligation to suspend judgement on all such matters. A more profound sceptic recognizes that thoroughgoing scepticism about rational justification actually allows for a level of doxastic promiscuity that is entirely compatible with possessing as rich a set of beliefs as would be possessed by any nonsceptical inquirer who is guided by observation and experiment rather than the fanciful products of an overheated imagination. We have seen, therefore, that radical scepticism of a near-global kind does not place its adherents under an obligation to practise near-universal suspension of judgement. It should be seen instead as merely exposing the hollowness of the supposed epistemic reasons that non-sceptics regard as constraining the beliefs they ought to hold. All other motives for holding beliefs remain potentially intact, although the balance of psychological forces that generate and sustain our beliefs is subtly rearranged. Beliefs founded on loose, unstable principles of association prove to be causally incapable of withstanding prolonged sceptical scrutiny, and the problems exposed within our epistemic practices by sceptical argumentation have the effect of reducing intellectual arrogance and dogmatism. It follows that if one is so constituted that one takes pleasure in complicated intellectual investigations, it is not the case that scepticism, even when accepted as true, provides any reason why one should puritanically refrain from such activities. In the words Hume uses as Book One of the Treatise approaches its

conclusion: ‘A true sceptic will be diffident of his philosophical doubts, as well as of his philosophical conviction; and will never refuse any innocent satisfaction, which offers itself, upon account of either of them.’ (THN 1.4.7.14 / 273) One’s willingness to accept that particular beliefs cannot be rationally justified is, moreover, entirely compatible with firmly regarding those beliefs as true and also having the meta-belief that these beliefs are the products of psychological mechanisms that mostly latch on to the truth in nearby possible worlds. Viewing beliefs as incapable of being supported by any form of reasoning or putative evidence that does not exhibit circularity or dependence on arbitrary and undefended assumptions is something that needs to be carefully distinguished from a stance of regarding those beliefs as arising from psychological mechanisms that would generate false beliefs in many possible worlds close to the one within which we take ourselves to exist. The first stance is one the Humean sceptic adopts in respect of almost all beliefs. The second stance, however, is potentially one with much more radical implications at a doxastic level. This latter stance is associated with either genuine suspension of judgement on non-epistemic matters or, in those cases where suspension of judgement fails to supervene, a state of alienation from one’s beliefs that mimics the attitude that a self-aware kleptomaniac or sufferer from a phobia takes towards the beliefs caused by his condition. Widespread suspension of judgement or doxastic alienation is not a phenomenon associated with the Humean sceptic, and it is the foregoing distinction that explains how this lack of disruption manages to co-exist with extremely wide-ranging scepticism about the availability of rational justification. It also

164

9780826443595_Ch06_Final_txt_print.indd 164

2/9/2012 9:35:15 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES serves to explain how the Humean sceptic can possess a high degree of confidence in the abilities of human beings to uncover the truth about even recondite matters of inquiry. The Humean sceptic views himself as fortunately located in an actual world that is one of those possible worlds in which many key mechanisms of belief formation latch on to the truth and also do this in the vast majority of neighbouring possible worlds. Thus he does not see his beliefs as the product of psychological mechanisms that generally fail to track the truth, and he is consequently neither alienated from those beliefs nor is he inclined to abandon them. At the same time he does accept that their truth is still, in some respects, a matter of unearned good fortune in the sense that it is impossible to construct any argument to defend the supposition that the actual world has the properties he ascribes to it that is not patently arbitrary or question begging. Thus he does regard all his beliefs, other than his H-minimal beliefs, as lacking rational justification.6 We arrive, therefore, at the combination of thoughts that gives rise to the distinctive character of Humean scepticism. The Humean sceptic is confident that he has investigative techniques at his disposal that can reliably yield more true beliefs than false beliefs. These are the techniques of ‘the experimental method of reasoning’, which is mentioned on the title pages of all three books of the Treatise. He also accepts, however, that any argument intended to show that this is the case must be probatively defective. Ultimately all that underpins this confidence is animal instinct and the positive psychological reinforcement we receive when our beliefs steer us away from pain or towards pleasure and also permit us to avoid encountering too many situations where we are driven to revise our previous beliefs.

It is one thing, however, to defend a habit of inquiry from criticisms that we should abandon this practice and to point out that such inquiry will inevitably lead us into momentary spasms of complete conviction, and quite another thing to maintain that such moments provide us with beliefs that can be assembled into a system that is beyond all criticism and challenge. Hume confesses, therefore, that although he has sometimes made use in the Treatise of ‘such terms as these, ’tis evident, ’tis certain, ’tis undeniable; which a due deference to the public ought, perhaps, to prevent’ (THN 1.4.7.15 / 274), these terms are to be understood merely as spontaneous expressions of momentary flashes of conviction that fail to retain their full vigour when placed in a broader context of countervailing considerations and general worries about the adequacy of human reason. Hume, accordingly, ends Book One of the Treatise with an observation that once again alerts us to Hume’s wish to fuse philosophical scepticism and the experimental method into one integrated and mutually supportive methodology for investigating both the natural world and human psychology. Hume declares that these emphatic declarations of certainty and conviction ‘were extorted from me by the present view of the object, and imply no dogmatical spirit, nor conceited idea of my own judgement, which are sentiments that I am sensible can become no body, and a sceptic still less than any other’ (ibid.). As Hume is referring here to characteristics of his own psychological constitution and denying that he possesses a ‘dogmatical spirit’ or ‘conceited idea of my own judgement’, there is surely no good motive for Hume to say that he is aware that these are dispositions that are not appropriate for a sceptic unless Hume is, at this point, thinking of himself as a sceptic.7 Thus Book One of the Treatise begins

165

9780826443595_Ch06_Final_txt_print.indd 165

2/9/2012 9:35:16 PM

HUME ON SCEPTICISM AND THE MORAL SCIENCES and ends with statements that are revelatory of the two central pillars of his overall philosophical outlook. As we have already stressed, Hume’s subtitle for the Treatise identifies it as seeking to apply the experimental method to moral subjects. Moreover, the final sentence of Book One constitutes Hume’s formal unmasking of himself as a sceptic. And the journey from that optimistic subtitle to the concluding announcement of his commitment to scepticism is very much an exploration of how scepticism is not only compatible with the successful application of the experimental method but also assists us to apply it correctly. As scepticism demolishes the pretensions of that method’s putative competitors and prevents us from being led astray by seemingly attractive lines of thought that are not genuinely grounded in the strong associative links that underpin compelling causal inferences, the Humean synthesis of radical scepticism and the experimental method represents a new and improved form of empiricism rather than the despairing renunciation of inquiry and reasoning that is so often attributed to scepticism by its detractors.

3

4

5

6

NOTES 1

2

In this sub-section of An Enquiry concerning Human Understanding, Hume uses the word ‘Pyrrhonism’ at 12.21 / 158 and the phrase ‘excessive scepticism’ at 12.23 / 159. ‘Pyrrhonian’ is used as a noun twice at 12.23 / 160, and the phrase ‘excessive principles of scepticism’ can be found at 12.21 / 158–9. The first detailed presentation of the case for interpreting Hume’s version of Academic scepticism as a causally rather than intellectually

7

mitigated form of scepticism seems to have been Robert Fogelin’s paper ‘The Tendency of Hume’s Skepticism’, in M. Burnyeat (ed.), The Skeptical Tradition (Berkeley, California: University of California Press, 1983), pp. 397–413. This paper dramatically improves our understanding of Hume’s scepticism by showing how we can avoid treating Hume’s Academic scepticism as a repudiation of the sceptical arguments found in the Treatise. Hume and his contemporaries used the term ‘natural philosophy’ to refer to what we now describe as the natural sciences, i.e. such disciplines as physics, chemistry and biology. The terms ‘moral science’ and ‘moral philosophy’ were used to refer to psychology and the social sciences along with various forms of intellectual activity that would still be collected without any disquiet under the banner of philosophy even today. Thus the attribution to Philo of ‘careless scepticism’ should not be seen as Hume’s way of insinuating some criticism of his views or his manner of presenting those views (DNR, Pamphilus, 128). The difference between standard regressive arguments for scepticism and Hume’s genetic arguments is usefully explored in R.J. Fogelin, ‘Hume’s Scepticism’, in D.F. Norton (ed.), The Cambridge Companion to Hume (Cambridge: Cambridge University Press, 1993), pp. 90–116. This characterization of the doxastic posture of the Humean sceptic draws heavily on Duncan Pritchard’s analysis of the distinction between veritic epistemic luck and reflective epistemic luck. See D. Pritchard, Epistemic Luck (Oxford: Oxford University Press, 2005), chap. 6, ‘Two Varieties of Epistemic Luck’, pp. 145–80. The important role played by the concluding sentence of Book 1 in the overall structure of the work is emphasized in D. Garrett, ‘“A Small Tincture of Pyrrhonism”: Skepticism and Naturalism in Hume’s Science of Man’, p. 90n2, in W. Sinnott-Armstrong (ed.), Pyrrhonian Skepticism (Oxford: Oxford University Press, 2004), pp. 68–98.

166

9780826443595_Ch06_Final_txt_print.indd 166

2/9/2012 9:35:16 PM

7 THE SELF AND PERSONAL IDENTITY Harold Noonan

1. INTRODUCTION Hume discusses the self and its identity in three main places: in sections V and VI of Part IV of Book I in the body of the Treatise of Human Nature and in an Appendix published a year later with Book III. In the last he declares himself wholly dissatisfied with his treatment of the topic in the body of the Treatise, but confesses that he now finds the whole matter a ‘labyrinth’ and identifies two principles he can neither render consistent nor renounce. There is no discussion of the topic in the Enquiry concerning Human Understanding. Unfortunately Hume fails to make clear in his recantation what he finds objectionable in his earlier account (since the principles he labels inconsistent are in fact consistent), and though commentators have produced a variety of suggestions, no consensus as to what Hume’s worry was has emerged. Section VI of Part IV of Book I of the Treatise is entitled ‘Of personal identity’. Our first task will be to identify exactly what this title denotes. The title of section V is ‘Of the immateriality of the soul’. However, this is somewhat misleading, since Hume’s main aim in this section is not to take sides but to criticize a presupposition of both sides of the debate between materialists and

immaterialists, namely, that the soul is a substance in which perceptions inhere. The attack on this proposition is the foundation of Hume’s discussion in section VI and the basis of his insistence that the self is a fiction. This is where we start.

2. THE FICTION OF PERSONAL IDENTITY In the tradition in which Hume was writing, deriving from Locke, the problem of personal identity was seen as that of giving an account of what constitutes personal identity. Locke’s own answer to this question has a negative component and a positive component. The negative component is that personal identity is not constituted by identity of substance, whether material or immaterial, any more than is identity of man: ‘it being one thing to be the same substance, another the same man, and a third the same person’.1 The positive component that does constitute personal identity is sameness of consciousness: ‘And as far as this consciousness can be extended backwards to any past action or thought, so far reaches the identity of that person’.2 Thus, Locke asserts, combining the two components, ‘it being the same consciousness

167

9780826443595_Ch07_Final_txt_print.indd 167

2/9/2012 9:35:27 PM

THE SELF AND PERSONAL IDENTITY that makes a man be himself to himself, personal identity depends on that only, whether it be annexed only to one individual substance or can be continued in a succession of several’.3 In subsequent discussions reacting to Locke the role of substance in the constitution of personal identity became the key issue and Butler, Reid and Leibniz all restored, in their accounts, the link that Locke had broken between personal identity and substantial identity.4 If we read Hume as contributing to this debate on the constitution of personal identity, we must understand his main contention to be an emphatic endorsement of the negative component of Locke’s account: personal identity is not constituted by identity of substance. But, in fact, to read Hume in this way is to misunderstand him. According to Hume, it is not merely that personal identity is not constituted by identity of substance; personal identity is a fiction; the ascription of identity over time to persons, a mistake (one we all necessarily make). There is no intelligible question of the form ‘In what does personal identity consist?’. There is only the genetic problem of specifying the psychological causes of the universal but mistaken belief in the existence of enduring persons. This is the problem Hume addresses in his discussion of personal identity. However, it is not, of course, in Hume’s view, a peculiarity of persons that they do not endure self-identically over time; nor does anything else. Hume thinks that the idea of identity is incompatible with the idea of change: it is the idea of an object which ‘remains invariable and uninterrupted thro’ a suppos’d variation of time’ (THN 1.4.6.6 / 253). This is Hume’s attempt to define an idea which is ‘a medium betwixt unity and number’ (THN 1.4.2.28 / 200), or ‘more properly speaking’, as he adds, ‘is either of

them, according to the view in which we take it’ (THN 1.4.2.29 / 201). Strictly speaking, it applies to nothing, since time requires change, but it certainly fails to apply to objects of ordinary discourse – plants, animals, artefacts and so on – and so, as in the case of persons, when we ascribe identity to them, Hume says, it is only in an ‘improper sense’. Thus, for Hume, the genetic problem of accounting for our false belief in the existence of enduring persons is just a part of the wider genetic problem of accounting for our false belief in the identity over time of changing things in general. In fact, he thinks, the same mechanism of the imagination which accounts for our ascriptions of identity over time to plants, animals and so on can equally well account for our ascriptions of identity over time to persons. This is because: The identity which we ascribe to the mind of man, is only a fictitious one and of a like kind with that which we ascribe to vegetable and animal bodies. It cannot, therefore, have a different origin, but must proceed from a like operation of the imagination upon like objects. (THN 1.4.6.15 / 259; emphasis added) The mechanism that generates the belief in the fiction of personal identity (the identity we ascribe to ‘the mind of man’) is the operation by which the mind is led to ascribe an identity to distinct but related perceptions, however interrupted or variable. It is a mechanism of the imagination in the narrow sense which is distinguished in the footnote to section IX of Part III of Book I: the word imagination is commonly used in two different senses; and though nothing be more contrary to true philosophy, than this inaccuracy, yet in the following

168

9780826443595_Ch07_Final_txt_print.indd 168

2/9/2012 9:35:27 PM

THE SELF AND PERSONAL IDENTITY reasonings, I have often been obliged to fall into it. When I oppose the imagination to the memory, I mean the faculty by which we form our fainter ideas. When I oppose it to reason, I mean the same faculty excluding only our demonstrative and probable reasonings. (THN 1.3.9.19 / 117)

we ascribe to plants and vegetables. And even when this does not take place, we will feel a propensity to confound these ideas, tho’ we are not fully able to satisfy ourselves in that particular, nor find anything invariable and uninterrupted to justify our notion of identity. (THN 1.4.6.6 / 254–5)

Hume has appealed to this identity-ascribing mechanism earlier in his account of the genesis of our belief in an external world or ‘body’ in section II and in his account of the ancient philosophers’ belief in material substrata in section III. As the distinction made in the quoted footnote indicates, he is emphatic that the beliefs generated by it are not thereby provided with a rational foundation, not even in the sense in which our beliefs in the unobserved effects or causes of observed causes or effects do have a rational foundation. Hume summarizes the effect of the action of the identity-ascribing mechanism of the narrow imagination as follows:

Hume indicates here the various applications of the identity-ascribing mechanism noted above: in addition to generating our belief in personal identity (i.e. a soul or self), it also generates our belief in substance (material substrata) as well as our belief in an external world or body, i.e. the continued existence of the perceptions of our senses, and our belief in the identity of plants and animals. The important point to note is that it is an essential element of this story, as Hume tells it, that the propensity we have to identify distinct perceptions is a propensity to regard them as identical in just the sense he defines: ‘an object that remains invariable and uninterrupted thro’ a supposed variation of time’. It is only if this is our idea of identity that the psychological mechanism can operate as he suggests. If we thought that identity were consistent with interruption or change then we would not be led to ‘feign the continued existence of the perceptions of our senses’ to remove the interruption or to ‘run into the notion of a soul, and self and substance’ or to be ‘apt to imagine something unknown and mysterious’ to disguise the variations. Hence it is essential to Hume’s account that our idea of identity is, in fact, the one he describes, and it is because this is so that he says:

In order to justify to ourselves this absurdity [i.e. the ascription of identity to distinct perceptions], we often feign some new and unintelligible principle, that connects the objects together, and prevents their interruption or variation. Thus we feign the continued existence of the perceptions of our senses to remove the interruption; and run into the notion of a soul, and self and substance, to disguise the variation, we may farther observe, that where we do not give rise to such a fiction, our propensity to confound identity with relation is so great, that we are apt to imagine something unknown and mysterious, connecting the parts, beside their relation; and this I take to be the case with regard to the identity

the controversy concerning identity is not merely a dispute of words. For when we attribute identity . . . to variable or interrupted objects, our mistake is not 169

9780826443595_Ch07_Final_txt_print.indd 169

2/9/2012 9:35:27 PM

THE SELF AND PERSONAL IDENTITY confined to the expression, but is commonly attended with a fiction, either of something invariable and uninterrupted, or of something mysterious and inexplicable, or at least with a propensity to such fictions. (THN 1.4.6.7 / 255)

3. THE REIFICATION OF PERCEPTIONS Although Hume’s account of identity makes it impossible for him to accept enduring selves and provides him with the starting point for his account of the genesis of our belief in personal identity, his conception of what the self would have to be if it existed is another ground for his rejection of it. One of the best known passages in Hume’s discussion of personal identity is his denial that he is introspectively aware of any self or mental substance: For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without a perception, and never can observe anything but the perception. (THN 1.4.6.2 / 252) Hume writes as if it is just a matter of fact that on looking into himself he fails to find anything but perceptions, but this sits ill with his emphatic denial that he has any idea of a self distinct from perceptions. I can be confident that I am not observing a tea-kettle now because I know what it would be like to be doing so. But if Hume has no idea of a self he presumably has no conception of what it would be like to observe one. In that case, however, how does he know that he is not doing so?

Another difficulty is that, as Chisholm puts it, it looks very much as though the self that Hume professes to be unable to find is the one that he finds to be stumbling – stumbling on to different perceptions.5 For Hume reports the results of his introspection in the first person: ‘I never catch myself without a perception’, ‘I never observe anything but the perception’. Hume’s denial is not, therefore, the straightforward empirical assertion it might at first appear to be. But then what is his basis for it? We must recall that Hume reifies perceptions. He starts from a conception of mental states according to which for a person to be in a mental state is for a certain relational statement to be true of that person: that he is perceiving a certain sort of perception (‘to hate, to love, to think, to feel, to see, all this is nothing but to perceive’ (THN 1.2.6.7 / 67). But if this is correct it is very natural that Hume should deny the introspective observability of the self. For if to be in any mental state is to possess a relational property, then no mental state can be an intrinsic quality of its subject. Given that the only states of which one can be introspectively aware are mental, then introspective awareness of a self would require awareness of it without any awareness of its intrinsic qualities. But surely it makes no sense to speak of observing something introspectively if the thing has no intrinsic qualities whatsoever which one can observe by introspection. These simple reflections suffice, I think, to explain Hume’s confidence in his denial of the introspective accessibility of the self. But they can be taken further if we now turn from what the Humean conception of the mental implies about the subject of mental states – that its only properties are relational ones – to what it implies about

170

9780826443595_Ch07_Final_txt_print.indd 170

2/9/2012 9:35:28 PM

THE SELF AND PERSONAL IDENTITY their objects, Hume’s perceptions. What it implies, of course, is that these perceptions are things, indeed substances, and they are logically capable of existing independently of being perceived. Hume is emphatic that this is the case. Indeed, he thinks that everything which can be conceived is a substance (THN 1.4.5.5 / 233). This is a consequence he explicitly draws from the conjunction of two of his fundamental principles in section V and makes him ‘absolutely condemn’ the question of the materiality or immateriality of the soul (THN 1.4.5.6 / 234). The first aim of section V is to put a stop to the endless cavils between ‘the curious reasoners concerning the material or immaterial substances in which they suppose our perceptions to inhere’ (THN 1.4.5.2 / 232). He first challenges these curious reasoners to explain what they mean by a substance and inhesion. He demands that they point out the impression from which their idea derives. And then he goes on to give his master-argument: If instead of answering these questions, anyone shou’d evade the difficulty, by saying, that the definition of substance is something which may exist by itself, and that this definition ought to satisfy us: Shou’d this be said, I shall observe that this definition agrees to everything, that can possibly be conceiv’d, and never will serve to distinguish substance from accident, or the soul from its perceptions. For thus I reason. Whatever is clearly conceiv’d may exist, and whatever is clearly conceiv’d after any manner may exist after the same manner. This is one principle . . . . Again, everything which is different is distinguishable, and everything which is distinguishable, is separable by the imagination. This is another principle. My conclusion from both is,

that since all our perceptions are different from each other, and from every thing else in the universe, they are also distinct and separable, and may be consider’d as separately existent, and may exist separately, and have no need of any thing else to support their existence. They are, therefore, substances, as far as this definition explains a substance. (THN 1.4.5.5 / 233)

This argument is perfectly general, of course, and if it is a good one it establishes that not only perceptions but qualities generally are logically capable of an independent existence. Hume is aware of this and indeed gives the argument himself in his criticism of the ancient philosophers’ conception of substance in section III: ‘every quality being a distinct thing from another, may be conceiv’d to exist apart, and may exist apart, not only from every other quality, but from that unintelligible chimera of substance’ (THN 1.4.3.7 / 222). But if so, Descartes’s famous analogy in the Second Meditation, in which he compares the relation between a piece of wax and its qualities to the relation between a man and his clothes, would be an appropriate one. But one consequence of this analogy is that the wax is represented as hidden beneath its garments and so as in itself unobservable. This is because the analogy implies that the assertion that the wax has any quality is in reality an assertion of a relation between it and something else. And a second consequence of the analogy is that the qualities of the wax are represented as being themselves substantial, as though they can ‘stand by themselves’, as a suit of armour can when no man is wearing it. But these consequences of the analogy, which is an appropriate one if the Humean argument is a good one, make it obvious that if the wax is so conceived, its existence, as

171

9780826443595_Ch07_Final_txt_print.indd 171

2/9/2012 9:35:28 PM

THE SELF AND PERSONAL IDENTITY universe of thought, or my impressions and ideas. There I observe another sun, moon and stars. . . . Upon my enquiring concerning these, theologians . . . tell me, that these also are modifications . . . of one simple substance. Immediately . . . I am deafen’d with the noise of a hundred voices, that treat the first hypothesis with detestation and scorn . . . and the second with applause and veneration . . . I turn my attention to these hypotheses . . . and find that they have the same fault of being unintelligible . . . and [are] so much alike, that . . . any absurdity in one . . . is . . . common to both. (THN 1.4.5.21 / 242–3; emphasis added)

anything other than that of a collection of qualities, must be regarded as highly problematic. Exactly the same is true of the self if Hume’s argument is correct.

4. THE REJECTION OF THE SUBSTANTIAL SELF With the argument of ‘Of the immateriality of the soul’ against substance behind us we can now turn to the details of Hume’s section on personal identity. But before doing so we should pause to look at another remarkable argument in section V which brings out how far Hume is prepared to proceed on the basis of his conception of perceptions as logically independent entities. Hume states that although he has already condemned as utterly unintelligible the question of the substance of the soul, he ‘cannot forbear proposing some further reflections’. Namely, that the doctrine of the immateriality, simplicity and indivisibility of a thinking substance is ‘a true atheism and will serve to justify all those sentiments for which Spinoza is so universally infamous’ (THN 1.4.5.17 / 240) and that this ‘hideous hypothesis’ [Spinozistic monism] is almost the same as that of the immateriality of the soul. Hume’s argument goes as follows:

Nor are matters improved for the theologians, according to Hume:

there are two different systems of beings presented, to which I suppose myself under a necessity of assigning some substance, or ground of inhesion. I observe first the universe of objects or of body: the sun, moon, stars, the earth . . . Here Spinoza . . . tells me that these are only modifications; and that the subject, in which they inhere, is simple, incompounded, and indivisible. After this I consider the other system of beings, viz. the

if instead of calling thought a modification of the soul, we should give it the more antient, and yet more modish name of an action. By an action we mean . . . something which, properly speaking, is neither distinguishable, nor separable from its substance. . . . But nothing is gained by this change of the term modification, for that of action. . . . First . . . the word action, according to this explication of it, can never be justly apply’d to any perception. . . . Our perceptions are all really different, and separable, and distinguishable from each other, and from every thing else. . . . [In] the second place . . . may not the Atheists likewise take possession of [the word “action”], and affirm that plants, animals, men, etc., are nothing but particular actions of one simple . . . substance? This . . . I own ’tis unintelligible but . . . assert . . . that ’tis impossible to discover any absurdity in the supposition . . . which will not be applicable to a like supposition concerning impressions and ideas. (THN 1.4.5.27–8 / 245–6)

172

9780826443595_Ch07_Final_txt_print.indd 172

2/9/2012 9:35:28 PM

THE SELF AND PERSONAL IDENTITY There could not, I think, be a clearer illustration than this of the seriousness with which Hume takes his reification of perceptions – if a tree cannot be a modification of Spinoza’s God, my idea of a tree cannot be a modification of me! Turning now to the section ‘Of personal identity’ Hume proceeds very rapidly, and confidently, for reasons that I hope will now be perfectly understandable, to his conclusion that the self is nothing more than a bundle of perceptions. The whole business takes less than two pages. Some philosophers have thought that ‘we are every moment intimately conscious of what we call our self’. But: ‘Unluckily all these positive assertions are contrary to that very experience which is pleaded for them, nor have we any idea of self, after the manner it is here explained, for from what impression could this idea be derived?’ Since the self is supposed to be an unchanging object, any impression of self must be constantly the same throughout the whole course of our lives. But, Hume finds, looking within himself, ‘There is no impression constant and invariable. Pain and pleasure, grief and joy . . . succeed each other . . . . It cannot therefore, be from any of these impressions, or from any other that the idea of self is deriv’d; and consequently there is no such idea’ (THN 1.4.6.2 / 251–2). Hume goes on to raise explicitly the difficulty that his conception of perceptions as ontologically independent creates for the notion of a substantial self: But farther, what must become of all our particular perceptions upon this hypothesis? All these are different, and distinguishable, and separable from each other, and may be separately consider’d, and may exist separately, and have no

need of any thing to support their existence. After what manner therefore do they belong to self; and how are they connected with it? (THN 1.4.6.3 / 252) It is immediately after this that he issues his denial of the observability of a self distinct from perceptions, and concludes that the self can be nothing but a bundle of perceptions. The same structure is exhibited in the Appendix, in which Hume summarizes his argument for the bundle theory before making his famous confession of bafflement. After arguing that we have no impression of self or substance as something simple or individual from which these ideas might be derived, he goes on to spend no less than three paragraphs insisting on the ontological independence of perceptions, finally concluding that since ‘’tis intelligible and consistent to say that objects exist distinct and independent, without any common single substance or subject of inhesion’ (that is, it is intelligible and consistent to reject Spinoza’s monism), ‘this proposition therefore can never be absurd with regard to perceptions’ (THN App. 14 / 634). In the immediately following paragraph he denies the observability of the self and derives the bundle theory. So much, then, for Hume’s arguments for the bundle theory of the self. Taken together with his analysis of identity, they entitle him, he believes, to the conclusion that personal identity is a fiction, that ‘the mind is a kind of theatre, where several perceptions successively make their appearance . . . . There is properly no simplicity in it at one time, nor identity in different’ (THN 1.4.6.4 / 253). For the idea of identity is that of an object, that ‘remains invariable and uninterrupted thro’ a suppos’d variation of time’. But if the bundle theory is correct, a person is nothing but a sequence of different (ontologically

173

9780826443595_Ch07_Final_txt_print.indd 173

2/9/2012 9:35:28 PM

THE SELF AND PERSONAL IDENTITY independent) objects existing in succession, and connected by a close relation – something like a thunderstorm. But ‘as such a succession answers perfectly to our notion of diversity, it can only be by [a] mistake that we ascribe to it an identity’ (THN 1.4.6.7 / 255). The only question that remains then, Hume thinks, is to explain the psychological mechanism that accounts for this mistake.

5. HUME’S ACCOUNT OF THE SOURCE OF THE MISTAKE Hume summarizes his account of this as follows. In contemplating an identical, that is, an invariable and unchanging object, we are doing something very different from contemplating a succession of objects related by links of resemblance, causation and contiguity but: That action of the imagination, by which we consider the uninterrupted and invariable object, and that by which we reflect on the succession of related objects, are almost the same to the feeling . . . . The relation facilitates the transmission of the mind from one object to another, and renders its passage as smooth as if it contemplated one continu’d object. This resemblance is the cause of the confusion and mistake, and makes us substitute the notion of identity, instead of that of related objects. However at one instant we may consider the related succession as variable or interrupted, we are sure the next to ascribe to it a perfect identity, and regard it as invariable and uninterrupted. (THN 1.4.6.6 / 254) This identity-ascribing mechanism of the imagination is also operative, as noted

previously, in generating our belief in an external world. Hume describes its operation here in ‘Of scepticism with regard to the senses’. According to Hume’s story, the constancy of our perceptions leads us to ascribe them a numerical identity, despite their interruptedness, since when we consider a constant and uninterrupted object, on the one hand, and reflect on an interrupted succession of constant objects, on the other, in the second situation there is the ‘same uninterrupted passage of the imagination’ (THN 1.4.2.34 / 203) as in the first. The second situation places the mind in the same ‘disposition and is considered with the same smooth and uninterrupted progress of the imagination, as attends the view of’ the first (THN 1.4.2.34 / 204). But ‘whatever ideas place the mind in the same disposition, or in similar ones, are apt to be confounded’ (THN 1.4.2.32 / 203). Thus I confound the two situations. But since I take the former to be a view of an identical object, I do the same with the latter situation and ‘confound the succession with the identity’ (THN 1.4.2.34 / 204). But Hume insists that it is an essential part of the notion of identity that an identical object must be uninterrupted as well as invariable in its existence. Thus I unite the ‘broken appearances’ by means of ‘the fiction of a continu’d existence’ (THN 1.4.2.36 / 205). That is, I come to believe that the identical perception which I earlier perceived has continued in existence whilst unperceived. Hume denies that this is a contradictory belief: ‘the continu’d existence of sensible objects or perceptions involves no contradiction’ (THN 1.4.2.40 / 208). Thus ‘we may easily indulge our inclination to that supposition’ (THN 1.4.2.40 / 208). In doing so, I reconcile my initially contradictory beliefs (in identity and interruptedness) by abandoning the second. I thus ‘remove entirely [the interruption] by

174

9780826443595_Ch07_Final_txt_print.indd 174

2/9/2012 9:35:28 PM

THE SELF AND PERSONAL IDENTITY supposing that these interrupted perceptions are connected by a real existence of which we are insensible’ (THN 1.4.2.24 /199), and arrive at a consistent set of beliefs. This is Hume’s view of the form that the belief in body takes in the minds of the vulgar, the non-philosophers: they believe that their very perceptions have a continued and distinct existence. This vulgar view involves no internal incoherence or concealed absurdity. It is a stable stopping place for the vulgar mind; as it were, a finished product of the imagination. But it is false. Merely as a matter of empirically discoverable fact, Hume thinks, perceptions are dependent and perishing existences. Experiments familiar to philosophers (but apparently not known to the vulgar) establish this (THN 1.4.2.45 / 211). But even philosophers cannot resist the force of the identity-ascribing mechanism of the imagination. They too must regard the second situation as the view of an identical object. However, they know that, as a matter of fact, perceptions do not continue unperceived. To resolve their conflict all they can do is to distinguish between objects and perceptions, ascribing continuity and distinctness to the former and interruptedness to the latter. But such a system of ‘double existence’, Hume thinks, is only a ‘palliative remedy’ and contains all the difficulties of the vulgar system, with others that are peculiar to itself (THN 1.4.2.46 / 211). The second operation of the identity-ascribing mechanism which Hume discusses before the section on personal identity is its operation in bringing about the ancient philosophers’ beliefs in material substrata. However, it operates differently in this case, in a way that causes Hume to say that whereas ‘we [i.e. the vulgar] feign the continu’d existence of the perceptions of our senses to remove the

interruption’ the ancient philosophers merely run into the notion of substance to ‘disguise the variation’ (THN 1.4.6.6 / 254). Hume explains the genesis of the belief in ‘substance, or original and first matter’ (THN 1.4.4.4 / 220) as follows: as the ideas of the several successive qualities of objects are united together by a very close relation, the mind, in looking along the succession . . . will not more perceive the change, than if it contemplated the same unchangeable object. But when we alter our method of considering the succession, and . . . survey at once any two distinct periods of its duration . . . the variations, which were insensible when they arose gradually, do now appear of consequence, and seem entirely to destroy the identity . . . . When we gradually follow an object in its successive changes, the smooth progress of the thought makes us ascribe an identity to the succession . . . . When we compare the succession after a considerable change, the progress of the thought is broke; and consequently we are presented with the idea of diversity: In order to reconcile which contradictions the imagination is apt to feign something unknown and invisible, which it supposes to continue the same under all these variations; and this unintelligible something it calls substance or original and first matter. (THN 1.4.4.3–4 / 220) There are three differences between this account of the genesis of the ancient philosophers’ belief in substance and Hume’s account in the previous section of our belief in body. First, no appeal is made to constancy. The belief is not the result of the effect on the mind of a constant but interrupted set of perceptions, but of a gradually and imperceptibly changing uninterrupted succession

175

9780826443595_Ch07_Final_txt_print.indd 175

2/9/2012 9:35:29 PM

THE SELF AND PERSONAL IDENTITY of qualities. Secondly, there is, therefore, no stable consistent stopping place akin to the vulgar belief in body, which is arrived at simply by endorsing the identity and rejecting the appearance of interruption in the sequence of perceptions as a mere appearance – the variation in the sequence of qualities cannot be rejected as a mere appearance and is inconsistent with the idea of identity in Hume’s account. Hence Hume does not distinguish in this case between the vulgar and philosophical views. There is only the philosophical view. The ancient philosophers’ belief is not a secondary product of the imagination, although it is a product of a multi-stage process. Thirdly, what presents us with the idea of diversity in this case, when the succession is viewed from a single point in time and two distinct periods of its duration surveyed, is not interruption, but variation. As noted, this cannot be regarded as a mere appearance, as interruption is in the vulgar view of body. So reconciliation of the two points of view is not possible and, as we have noted, Hume carefully speaks only of ‘disguising’ the variation to reconcile the contradiction. The ancient philosophers’ belief in substance is not a consistent one since it involves the idea of something which is both identical and changing. The identity-ascribing mechanism of the narrow imagination works to produce conflation and error in exactly the same way, Hume thinks, in the case of personal identity. The succession of my perceptions is merely a succession of distinct related objects. But because the objects in the succession are closely related, the action of the imagination in surveying the succession is ‘almost the same to the feeling’ as the action of the imagination in considering an uninterrupted and invariable object. As in the other cases, the similarity between the two acts of mind leads me to confound the two situations and

thus to regard the succession of related perceptions as really united by identity. And so I am led to believe in the unity of the self, which is as much a fiction as in the other cases of the operation of the identity-ascribing mechanism, and, ‘proceed[s] entirely from the smooth and uninterrupted progress of the thought along a train of connected ideas according to the principles above explain’d’ (THN 1.4.6.16 / 260). All that remains to be said, Hume thinks, is what the relations are in this case that link my successive perceptions so as to bring about this uninterrupted progress of the thought. His answer is: resemblance and causation. Our perceptions at successive times resemble each other for a variety of reasons, of course, but the one Hume emphasizes is that people can remember their past experience: For what is the memory, but a faculty by which we raise up the images of past perceptions? And as an image necessarily resembles its object must not the frequent placing of these resembling perceptions in the chain of thought, convey the imagination more easily from one link to another, and make the whole like the continuance of one object? (THN 1.4.6.18 / 260–1) Given this copy theory of memory Hume is able to regard memory not merely as providing us with access to our past selves, but also as contributing to the bundles of perceptions which we can survey, elements which represent, and thus resemble, earlier elements; and so – since resemblance is a relation which enables the mind to slide smoothly along a succession of perceptions – as strengthening our propensity to believe in the fiction of a continuing self. In this particular case, then, Hume is able to say, with a nod of agreement

176

9780826443595_Ch07_Final_txt_print.indd 176

2/9/2012 9:35:29 PM

THE SELF AND PERSONAL IDENTITY to Locke, ‘memory not only discovers the identity but contributes to its production’ (THN 1.4.6.18 / 261). But we do not remember all, or even most of, our past actions or experiences. Yet we do not affirm, because we have entirely forgotten the incidents of certain past days, that the present self is not the same person as the self of that time. Consequently there must be something else which enables us to think of our identity as extending beyond our memory. Here Hume appeals to causality, which has been previously introduced in his account of:

According to this account of how belief in personal identity arises via the identityascribing mechanism of the imagination, it operates as it does in producing the ancient philosophers’ belief in material substrata and not as it does in producing belief in body, as Hume indicates: ‘we feign the continu’d existence of the perceptions of our senses to remove the interruption; and run into the notion of a soul, and self, and substance to disguise the variation’ (THN 1.4.6.6 / 254). Our belief in personal identity is not explained by the constancy of our perceptions. (On the contrary, these:

the true idea of the human mind . . . a system of different perceptions or different existences, which are linked together by the relation of cause and effect . . . . In this respect I cannot compare the soul more properly to anything than to a republic or commonwealth, in which the several members are united by the reciprocal ties of government and subordination, and give rise to other persons, who propagate the same republic in the incessant changes of its parts. (THN 1.4.6.17 / 261)

succeed each other with an inconceivable rapidity, and are in perpetual flux and movement. Our eyes cannot turn in their sockets without varying our perceptions. Our thought is still more variable . . . , and all our other senses and faculties contribute to this change; nor is there any single power of the soul, which remains unalterably the same perhaps for one moment (THN 1.4.6.4 / 253).)

When we think of ourselves as existing at times we cannot remember we do so, Hume says, by imagining the chain of causes and effects that we remember extending beyond our memory of them. So the causal links between our perceptions, as well as their resemblances, are crucial to our belief in a continuing self which exists at times it no longer recalls. Consequently, Hume is able to say, this time in agreement with Locke’s opponents: ‘In this view . . . memory does not so much produce as discover personal identity, by shewing us the relation of cause and effect among our different perceptions’ (THN 1.4.6.20 / 262).

And, like the ancient philosophers’ belief, it is not in fact a successful attempt at reconciliation of our belief in identity with the variation we cannot deny, but merely an attempt to disguise the variation. This is why there is no distinction between vulgar and philosophical forms of the belief in personal identity. In fact, what disposes us to regard temporally distinct and sensibly distinguishable (non-resembling) perceptions as identical is primarily their causal relatedness. This is present when perceptions are neither invariable nor uninterrupted, that is, when neither of the conditions specified in Hume’s account of our idea of identity is satisfied. Nevertheless, Hume thinks, just as the ancient philosophers were caused by the imperceptibility of the differences between

177

9780826443595_Ch07_Final_txt_print.indd 177

2/9/2012 9:35:29 PM

THE SELF AND PERSONAL IDENTITY successive qualities to regard them as identical and thence were led to the notion of substance in order to reconcile this with the perceptible differences existing when the succession was viewed from a single point of time, so we are caused by the causal relatedness of our perceptions to regard them as identical and fall into the notion of a soul or self in an attempt to reconcile this with their variation, which makes their non-identity evident. (Of course, no one ever thinks that distinguishable perceptions are identical, but it is our inclination to think this, together with our recognition that it cannot be so which, according to Hume, explains our belief in a self.) Thus, unlike our vulgar or everyday belief in body, our everyday belief in a self is not a consistent, stable, and merely contingently false one, which can only be revealed to be false by reflection on experiments to which the non-philosophical (in particular, the ‘honest gentlemen’ of England (THN 1.4.7.13 / 272)) may not be inclined. Like the ancient philosophers’ belief in material substrata, it is an inevitably doomed attempt to reconcile the irreconcilable. It needs no argument that Hume’s supplementary account of our belief in the simplicity of the self (THN 1.4.6.23 / 263) is to be compared to his account of the ancient philosophers’ belief in the simplicity of material substrata since it is just a summary of the account of that given in section III (THN 1.4.3.5 / 221).

in every system concerning external objects, and the idea of matter, ‘the intellectual world . . . is not perplexed with any such contradictions as those we have discovered in the natural’ (THN 1.4.5.1 / 232). By the time of the Appendix this confidence has disappeared. The topic of personal identity is now declared a ‘labyrinth’, and two principles are identified which he can neither render consistent nor renounce, viz. that all our distinct perceptions are distinct existences, and that the mind never perceives any real connection among distinct existences. Hume’s identification of his difficulty is brief and inaccurate (since the principles he labels inconsistent are in fact consistent). The most a commentator can do is to identify a flaw in his discussion of personal identity of which he plausibly could have become aware after completing Book I and which it is consistent with the discussion in the Appendix to suppose he might have had in mind as the cause of the trouble. It appears that what worries Hume is some defectiveness in his explanation of our belief in personal identity, and specifically in his explanation of our initial tendency to attribute identity to what are in fact distinct perceptions: Having thus loosen’d all our particular perceptions, when I proceed to explain the principle of connexion, which binds them together, and makes us attribute to them a real simplicity and identity, I am sensible, that my account is very defective, and that nothing but the seeming evidence of the precedent reasonings cou’d have induced me to receive it. (THN App. 20 / 635)

6. THE LABYRINTH In Book I Hume is confident that despite the contradictions and difficulties that exist

Again: ‘But all my hopes vanish, when I come to explain the principles that unite our

178

9780826443595_Ch07_Final_txt_print.indd 178

2/9/2012 9:35:29 PM

THE SELF AND PERSONAL IDENTITY successive perceptions in our thought and consciousness. I cannot discover any theory, which gives me satisfaction on this head.’ (THN App. 20 / 635) As to Hume’s talk of inconsistency: the obviously consistent principles he declared to be inconsistent, but ones he cannot renounce, obviously entail that the mind never perceives any real connection among its perceptions. Presumably Hume thinks this inconsistent with the possibility of explaining our attribution of identity to distinct perceptions. Taking this focus on explanation into account, and by the standards indicated above for the satisfactoriness of an interpretation of Hume’s discussion, a suggestion as reasonable as any other is that Hume has here come to realize that in this case his appeal to the identity-ascribing mechanism cannot explain the irresistible and undeniable belief in personal identity he thinks we all have. As we have seen, our belief in an enduring self, in Hume’s account, is on a par with the ancient philosophers’ belief in substrata. Like the latter, then, it should easily be subverted by a due contrast and opposition (THN 1.4.4.1 / 225). In fact, in Hume’s account our belief in personal identity should be even less secure than the belief in material substrata. For, as Hume emphasizes, in this case the crucial relation between perceptions which causes us to pass easily from one to another, and consequently to attribute to them an identity, is cause and effect. But nowhere else are we disposed to identify cause and effect – even if we are disposed to add a new relation to objects perceived as related by cause and effect as a result of our disposition to complete the union (THN 1.4.5.12 / 237), we are not disposed to add the relation of

identity to causes and effects we perceive as distinct in space and time and non-resembling, that is, as possessing neither of the crucial qualities in his idea of identity. In his attempt to prove that ‘all objects, to which we ascribe identity, without observing their invariableness, and uninterruptedness, are such as to consist of a succession of related objects’ (THN 1.4.6.7 / 255), Hume identifies various special features of such successions which dispose the mind to a smooth and easy transition – a small and inconsiderable change in a part in proportion to the whole, gradual and insensible change, a combination of parts to a common end, and a sympathy or reciprocal relation of parts, a specific identity or resemblance and a change that is natural and essential and hence expected. But he makes no attempt at all to show that such features are present in the succession of different perceptions which ‘succeed each other with an inconceivable rapidity and are in a perpetual flux and movement’ (THN 1.4.6.4 / 252) that we think of as ourselves. He simply insists that, since the true idea of the human mind is to consider it as a system of different perceptions or different existences linked together by the relation of cause and effect, our ascription of identity to our distinct perceptions must be caused by the ease of transition between causally related perceptions. But this is merely an attempt to shoehorn an account of our belief in personal identity into the explanatory framework he has appealed to in Sections II and III of Part IV of Book I. I suggest that in the Appendix Hume has come to recognize the implausibility of this attempt and specifically, of its necessary first stage, the explanation of our disposition to regard distinct perceptions as themselves identical.

179

9780826443595_Ch07_Final_txt_print.indd 179

2/9/2012 9:35:30 PM

THE SELF AND PERSONAL IDENTITY NOTES 1

2 3 4

J. Locke, An Essay concerning Human Understanding, ed. J. Yolton (London: Dent, 1961), II.xxvii.7. Locke, Essay, II.xxvii.9. Ibid. See J.Butler, ‘Of Personal Identity’, First Dissertation to the Analogy of Religion [1736], repr.

5

in A. Flew (ed.), Body, Mind and Death (New York: Macmillan, 1964); T. Reid, Essays on the Intellectual Powers of Man, ed. A.D. Woozley (London: Macmillan, 1941); G.W. Leibniz, New Essays on Human Understanding, trans. and ed. P. Remnant and J. Bennett (Cambridge: Cambridge University Press, 1981). R.M. Chisholm, Person and Object (London: Allen and Unwin, 1976), p. 39.

180

9780826443595_Ch07_Final_txt_print.indd 180

2/9/2012 9:35:30 PM

8 ‘ALL MY HOPES VANISH’: HUME ON THE MIND Galen Strawson

1. ‘THE ESSENCE OF THE MIND [IS] UNKNOWN’ Hume holds that ‘the essence . . . of external bodies’ is unknown, and that ‘the essence of the mind [is] equally unknown to us with that of external bodies’ (THN Intro. 8 / xvii). His aim in section 1.4.6 / 251–63 of his Treatise, which addresses the question of the nature of the mind (after having dismissed the traditional debate between the materialists and immaterialists in the preceding section), is accordingly modest. It is, first, to provide an account of (1) the content of the empirically warranted idea of the mind (or self or person), given that we cannot know the essence of the mind, and, second, to provide a causal psychological account of the origin of (2) our belief in a single diachronically persisting mind (or self or person), given that the account of (1) turns out to show that no idea of the mind as a single diachronically persisting entity is empirically warranted.1 More particularly, it is to provide a psychological account of how each of us individually comes to

believe that he or she is a single persisting person or self or has a single persisting mind. When Hume returns to this topic in the Appendix to the Treatise (THN App. 10–21 / 633–6), he can find no fault in his accounts of (1) and (2). The consequence, in his own words, is that his ‘hopes vanish’. Why is this? He remains entirely happy with his account of (2), and reaffirms it in the Appendix (THN App. 20 / 635).2 The trouble lies in his account of (1). But the trouble is not that the account of (1) is wrong. On the contrary: Hume cannot see how it can be wrong, on his empiricist terms, and he reaffirms it, too, in the Appendix (THN App. 15–19 / 634–5). The trouble is that his philosophy as a whole makes essential use of a conception of mind that is not empirically warranted, according to (1). Since he is committed to an empiricist approach, he needs to get more into his account of the empirically warranted idea of the mind. But he cannot. So his hopes vanish.

2. THE EMPIRICALLY WARRANTED IDEA OF THE MIND What is the empirically warranted idea of the mind, the account of the mind at which any

181

9780826443595_Ch08_Final_txt_print.indd 181

2/9/2012 9:35:42 PM

HUME ON THE MIND philosophy that aims at clarity and distinctness must aim, given that the essence of the mind is unknown?3 Hume’s answer is plain: the mind ‘as far as we can conceive it, is nothing but a system or train of different perceptions’ (Abs. 28 / 657). We have no ‘notion of . . . self . . . , when conceiv’d distinct from particular perceptions . . . we have no notion of . . . the mind . . . , distinct from the particular perceptions’ (THN App. 18–19 / 635).4 These are explicitly epistemologically qualified statements of what has come to be known as ‘the bundle theory of mind’. They are claims to the effect that this is all we can know of the mind. But we also find many (mostly earlier) epistemologically unqualified ontological formulations of the bundle theory of mind. Minds or selves or persons or thinking beings, Hume says, are ‘nothing but a bundle or collection of different perceptions’ (THN 1.4.6.4 / 252). ‘They are the successive perceptions only, that constitute the mind’ (THN 1.4.6.4 / 253). A ‘succession of perceptions . . . constitutes [a] mind or thinking principle’ (THN 1.4.6.18 / 260). It is a ‘chain of causes and effects, which constitute our self or person’ (THN 1.4.6.20 / 262).5 A ‘composition of . . . perceptions . . . forms the self’ (THN App. 15 / 634). A ‘train . . . of . . . perceptions . . . compose a mind’ (THN App. 20 / 635). Hume could not be more plain: ‘what we call a mind, is nothing but a heap or collection of different perceptions, united together by certain relations’ (THN 1.4.2.39 / 207). It is a ‘connected mass of perceptions, which constitute a thinking being’ (THN 1.4.2.39 / 207), ‘a connected heap of perceptions’ (THN 1.4.2.40 / 207). It is a ‘succession of perceptions, which constitutes our self or person’ (THN 1.4.7.3 / 265). ‘It must be our several particular perceptions, that compose the mind. I say, compose the mind, not belong to it’ (Abs. 28 / 658).6

How should we take these remarks? Well, Hume holds that ‘the perceptions of the mind are perfectly known’ (THN 2.2.6.2 / 366), and he also holds that the essence of the mind is unknown. So he does not intend these outright ontological claims to be without restriction, as stating the essence of the mind. They are claims about the mind so far as we have any empirically – and hence philosophically – respectable knowledge of it.7 They are claims about the maximum legitimate content of any claims about the nature of the mind that can claim to express knowledge of the nature of the mind. Hume is a sceptic, and a sceptic, even a moderate sceptic like him, does not go around claiming to have certain a posteriori knowledge of the ultimate metaphysical nature of the concrete constituents of the universe (other than perceptions). He does not claim to have certain a posteriori knowledge either of the essence of the mind, or of the essence of objects. – The outright ontological claims about the mind are literally true when they are made strictly within the philosophical framework of ideas constituted by empirically warranted (and hence clear and distinct) ideas, since they just repeat the definition of the empirically warranted clear and distinct idea of the mind. True. The point is then that this framework of empirically warranted (and hence clear and distinct) ideas is in Hume’s philosophy rightly and crucially embedded within a larger sceptical framework of ideas. In the larger sceptical framework of ideas it is acknowledged that there may be and indeed is more to reality than what we can comprehend in empirically warranted ideas, and words like ‘mind’ and ‘bodies’ are accordingly used in a larger sense: ‘the essence of the mind’ is ‘equally unknown

182

9780826443595_Ch08_Final_txt_print.indd 182

2/9/2012 9:35:42 PM

HUME ON THE MIND to us with that of external bodies’ (THN Intro. 8 / xvii), and ‘the essence and composition of external bodies are so obscure, that we must necessarily, in our reasonings, or rather conjectures concerning them, involve ourselves in contradictions and absurdities’ (THN 2.2.6.2 / 366).

3. WHY IS THE EMPIRICALLY WARRANTED IDEA OF THE MIND NOT ENOUGH FOR HUME? Why is the empirically warranted idea of the mind not enough for Hume? Before trying to answer this question, we should consider Hume’s statement of what he would need in order to be able to put things right, now that his hopes have vanished. He is very straightforward about this: ‘Did our perceptions either inhere in something simple and individual, or did the mind perceive some real connexion among them, there wou’d be no difficulty in the case’ (THN App. 21 / 636). If we could appeal to the idea that our perceptions inhered in something simple and individual, all would be well. But in that case we would of course have to have empirical warrant for the view that our perceptions inhere in something simple and individual, and that, Hume thinks (rightly, on his terms), is something we will never have. We would also be able to put things right if we could perceive ‘some real connexion’ among our perceptions. For in that case the idea of real connection among perceptions would be empirically warranted (because perceived), and could accordingly feature as part of the empirically warranted idea of the mind. But that, Hume says (again rightly, on his terms), is something that will never happen.

4. WHAT DOES HUME NEED TO DO THAT HE CANNOT DO? If either of the two options specified in the previous quotation were available, Hume would be all right, given what he needs to do. What does he need to do? He is very straightforward about this too. He needs to explain something. What does he need to explain? He states what he needs to explain in two different ways in a single paragraph (THN App. 20 / 635–6). It is slightly confusing, because he speaks of principles first in the singular and then in the plural, although he has the same thing in mind in both cases. He needs ‘to explain the principle of connexion, which binds [our perceptions] together, and makes us attribute to them a real simplicity and identity’ (THN App. 20 / 635). He needs in other words ‘to explain the principles, that unite our successive perceptions in our thought or consciousness’ (THN App. 20 / 636). One finds exactly the same shift between the singular and the plural in the passage about the mind (in particular the ‘imagination’) in which his problem originates, and to which one must turn first when trying to say what his problem was:8 nothing wou’d be more unaccountable than the operations of [the imagination], were it not guided by some universal principles, which render it, in some measure, uniform with itself in all times and places. Were ideas entirely loose and unconnected, chance alone wou’d join them; and ’tis impossible the same simple ideas should fall regularly into complex ones (as they commonly do) without some bond of union among them, some associating quality, by which one idea naturally introduces another. This uniting principle among ideas . . . (THN 1.1.4.1 / 10–11)

183

9780826443595_Ch08_Final_txt_print.indd 183

2/9/2012 9:35:43 PM

HUME ON THE MIND Why does he need to explain this principle, or these principles, and what exactly does he mean by ‘explain’? At the very least, he means that he needs to make the existence of these principles readily intelligible. Why so? Well, for one thing, they are his fundamental explanatory posits. The fact that the mind is governed by these principles – the fact that the mind is a Principle-Governed Mind – is the fundamental explanatory posit of Hume’s whole philosophy. He needs to explain the Principle-Governed Mind – to make its existence readily intelligible. As a sceptic, Hume is clear on the point that he does not have to explain everything. He is clear that many things about the nature of the universe lie beyond the reach of human understanding. In fact he is clear on the point that he does not have to explain everything about the very principles that are in question when his hopes vanish. When we consider these principles, he says, ‘the principles of union or cohesion among our simple ideas’ (THN 1.1.4.6 / 12), we encounter a kind of attraction, which in the mental world will be found to have as extraordinary effects as in the natural . . . .[9] Its effects are every where conspicuous; but as to its causes, they are mostly unknown, and must be resolv’d into original qualities of human nature, which I pretend not to explain.

that case his enquiry wou’d be much better employ’d in examining the effects than the causes of his principle. (THN 1.1.4.6 / 12–13)10 Quite so. This is clear. But Hume is no less clear on the fact that there is something about these principles, about the PrincipleGoverned Mind, that he needs to explain and cannot explain; something that he needs to make readily intelligible, and cannot.

5. WHAT HAS CAUSED HIS DIFFICULTY TO ARISE IN THE FIRST PLACE? What has caused his difficulty? Hume is clear on this point too. It is the fact that he has ‘loosen’d all our particular perceptions’: ‘But having thus loosen’d all our particular perceptions, when11 I proceed to explain the principle of connexion, which binds them together, . . . I am sensible, that my account is very defective . . .’ (THN App. 20 / 635). Why is this loosening a problem? Because, he says, in a passage already quoted: were ideas entirely loose and unconnected, chance alone wou’d join them; and ’tis impossible the same simple ideas should fall regularly into complex ones (as they commonly do) without some bond of union among them, some associating quality, by which one idea naturally introduces another. (THN 1.1.4.1 / 10)12

He continues: Nothing is more requisite for a true philosopher, than to restrain the intemperate desire of searching into causes, and having establish’d any doctrine upon a sufficient number of experiments, rest contented with that, when he sees a farther examination would lead him into obscure and uncertain speculations. In

It seems, then, that he should not have loosened all our perceptions. If he had not, he would not have this problem. But when he asks what an empirically warranted idea of the mind looks like, he finds he has to loosen them. For all that is observable of the mind is a 184

9780826443595_Ch08_Final_txt_print.indd 184

2/9/2012 9:35:43 PM

HUME ON THE MIND ‘train of perceptions’ with no observable connection between them. This is his problem. In another well-known passage, Hume writes that ‘the true idea of the human mind, is to consider it as a system of different perceptions or different existences, which are link’d together by the relation of cause and effect and mutually produce, destroy, influence, and modify each other’ (THN 1.4.6.19 / 261). Does he not here claim that he has the ‘real connexion’ he needs? No. This reference to the relation of cause and effect brings in nothing more than the empirically warranted idea of cause and effect, and this, of course, is not an idea of real connection.13 Hume is clear on the point. The ‘true’, i.e. empirically warranted, idea of the human mind, which is characterized in terms of causal links, as above, is an account of the mind in which ‘all our particular perceptions’ are ‘loosen’d’ (THN App. 20 / 635). He cannot explain the existence of the principle/principles whose existence needs to be explained if all he has are loose perceptions – which are all his empirically warranted account of the mind gives him. He needs observable ‘real connexion’ between the perceptions, or their inherence in ‘something simple and individual’, which is in effect just a particularly strong form of real connexion (THN App. 21 / 636). That is what he says. He could hardly be more clear on the point.

of how it is that we come to believe we experience causal necessity or power in the world, and of how it is that we do this in spite of the fact that the actual legitimate empirical content of our experience of causation in the world contains no experience of causal necessity or power in the world. He uses this conception of the Principle-Governed Mind again in THN 1.4.2 / 187–218 (‘Of scepticism with regard to the senses’), and with equal success, in his psychological account of how it is that we come to believe in external objects that continue to exist independent and unperceived, and of how it is that we come to do this in spite of the fact that this property of continuous independent unperceived existence cannot be part of our experience, or, therefore, part of the actual legitimate empirical content of our idea of an external object. His conception of the imagination – I will capitalize the term ‘Imagination’ to mark it as Hume’s theoretical term – effectively contains his whole conception of the Principle-Governed Mind. It drives his account of how, given only 3. the actual empirically warranted content of the idea of causation, we none the less acquire 4. our belief that we experience causal power in external objects, as just remarked, and, equally, his account of how, given only 5. the actual empirically warranted content of the idea of an external object,

6. THE HEART OF THE PROBLEM We can put Hume’s problem in a suitably painful way as follows. He uses a certain theoretical conception of the mind – the Principle-Governed Mind described in THN 1.1.4 / 10–13 – with tremendous success in THN 1.3.14 / 155–72 (‘Of the idea of necessary connexion’), in his psychological account

we none the less acquire 6. our belief in external objects. And so far all is well and good.14 But when he comes (in THN 1.4.6 / 251–63) to give his account of how we acquire (2) our belief in a single diachronically persisting mind, given

185

9780826443595_Ch08_Final_txt_print.indd 185

2/9/2012 9:35:43 PM

HUME ON THE MIND only (1) the actual empirically warranted idea of the mind, he finds that (1), the actual empirically warranted idea of the mind, does not contain the materials – the machinery, one might say – he needs to drive his account of how we acquire (2). Nor does it contain the machinery he needs to drive his accounts of how we acquire (4) and (6). His account of (1), the actual empirically warranted idea of the mind, the only one he can use in his philosophy, has deprived him of the PrincipleGoverned Mind, the Imagination, which drives all the most original parts of his philosophy: the accounts of how we acquire (4), (6) and (2). The Imagination cannot have any real existence if all there is to the mind is a bundle of perceptions. There is nowhere for it to be. More moderately, and more precisely to the point: we cannot make use of the notion of the Imagination or the Principle-Governed Mind in our philosophy, if our philosophy rules that the only idea of the mind that is suitably clear and distinct, and can therefore be legitimately used in philosophy to make knowledge claims, is the bundle theory of mind. No doubt what we need exists. It exists in reality – it is the ‘essence of the mind’. But ‘the essence of the mind [is] unknown’, and we cannot make use of it in our empiricist philosophy. In sum: when Hume comes to give his empiricist account of the mind, his philosophy shoots itself in the foot. That is why his hopes vanish. His particular brand of empiricism is unsustainable.15 In the first Enquiry he endorses the claim that prompts his confession of failure in the Appendix to the Treatise, when, in his only direct reference to the abandoned problem, he observes, briefly but decisively, that ‘it is evident that there is a principle of connexion between the different thoughts or ideas of the mind’ (EHU 23 / 3.1). Here he refers to a real connection of precisely the sort that

the empiricistically ‘true’ idea of the human mind cannot countenance.16 This is the very same ‘principle of connexion’ that he refers to in the Appendix, the principle of connexion that, he says, makes ‘my hopes vanish . . . when I proceed to explain’ it (THN App. 20 / 635). He cannot explain it in the terms of his empiricist theory of ideas. In turning his attention to our idea of the mind in THN 1.4.6 / 251–63 – after having devoted detailed attention to our idea of causation in 1.3.14 / 155–72, and our idea of external bodies in 1.4.2 / 187–218 – Hume finds himself obliged to deprive himself of ‘the uniting principle’ (THN 1.1.4.1 / 10) that he had relied on in 1.3.14 and 1.4.2: ‘the principle of connexion, which binds . . . all our particular perceptions together’ (THN App. 20 / 636). It cannot be part of the empirically warranted idea of the mind. It cannot be part of the empirically warranted idea of the mind whether it is thought of as a ‘real connexion’ of some sort or as something ‘simple and individual’, because neither of these things is given in experience.17

7. THE UNANSWERABLE OBJECTION? Some may object that Hume does not really have the problem he thinks he has. I think he is not so foolish as to mistake his own situation, but the objection is worth addressing, and I will do so in due course. First, I will restate the issue. We can start with the two fundamental theoretical principles that Hume says he cannot renounce (THN App. 21 / 636). I will number them ‘(P1)’ and ‘(P2)’: (P1) ‘all our distinct perceptions are distinct existences’ (so far as we know)

186

9780826443595_Ch08_Final_txt_print.indd 186

2/9/2012 9:35:43 PM

HUME ON THE MIND and (P2) ‘the mind never perceives any real connexion among distinct existences’. These principles are very clear and familiar to any reader of Hume. In the next sentence he states two options, already noted, either of which would in his opinion entirely solve his problem. The first is that (O1) ‘our perceptions . . . inhere in something simple and individual’. The second, to spell it out a little, is that (O2) our perceptions are distinct existences, and ‘the mind perceive[s] some real connexion among them’. Why would either of these two options solve his problem? Because both can sufficiently ground a fundamental theoretical commitment of his philosophy – his commitment to the real existence and operation of something he has just mentioned,

of view. The ideas of real or objective unity or connection that the options appeal to are conceptually clear. They are ‘perfectly distinct’ as far as they go.18 But they are empirically ‘unintelligible’ in their application to concrete reality. So Hume cannot appeal to them in his account of the nature of the mind. His hopes vanish. This may not be the best way to put the point. It may be better to say that what happens in the Appendix is that Hume realizes that he has no answer, given his overall theory, to an objection that begins by citing the two options he mentions and then challenges him to deny that at least one of them (or perhaps their disjunction) is in effect built into what he means by ‘mind’, and is therefore built into what he is really taking to be the ‘true idea’ of the mind, although they are excluded from his official account of the ‘true idea’ of the mind. According to this account, Hume realizes that he faces what one might call the Unanswerable Objection:

(P3) the principle of connexion, which binds . . . our particular perceptions . . . together (THN App. 20 / 635), or (in its plural version) (P4) the principles, that unite our successive perceptions in our thought or consciousness (THN App. 20 / 636), most centrally, the principles of the association of ideas, the Resemblance principle, the Contiguity principle, and the Cause and Effect principle. The trouble is that both options are ruled out for him. They have no empirical warrant. They are philosophically inadmissible from his empiricist point

– Your philosophy taken as a whole commits you, Hume, to a view forbidden by your philosophy. More precisely, it commits you to a choice between one of two views. The first is that the self or mind is something like an ontologically (substantially) simple and individual persisting something, in which successive perceptions (i.e. experiences) inhere in such a way that it is not problematic that they are connected or united in the way you take them to be. If you reject this – as you must on your empiricist principles, given that we have (among other things) no warrant for believing in or appealing to the existence of anything that lasts longer than the duration of a single fleeting perception – and hold instead that the mind is something ontologically (substantially) multiple, you are no better off. For then you must hold

187

9780826443595_Ch08_Final_txt_print.indd 187

2/9/2012 9:35:43 PM

HUME ON THE MIND that the mind is something whose existence involves ‘real connexion’ (and this is something that you have in effect already done), and also, crucially, given your own fundamental empiricist principles, that this real connection is empirically knowable, experienceable or perceivable by us. You must hold this because the idea of real connection is built into the conception of mind you make use of in your philosophy, so it must be empirically justified in order to be licensed for use. But you must also reject the view that it is empirically knowable, given those same fundamental empiricist principles, and you have indeed done so.

way that it can be appealed to in empiricist philosophy’.20 Some students of Hume have difficulty with the idea that he makes any use at all of the idea of objective or real connection. This is understandable, at least at first, but Hume’s position is clear: if (once again) our ideas were

It seems that Hume agrees. He puts forward exactly the same two metaphysical options, (O1) and (O2), one straightforwardly ontological, the other ontological/epistemological, and then says that either would solve his problem, but that he cannot have either. It may be protested that Hume cannot really be saying this, because he takes the idea of a persisting, simple and individual something and the idea of real connection to be ‘unintelligible’ tout court. But he is saying this. Even those who want to reject the quotation in note 18 from THN 1.4.6.6 / 253, in support of the claim that Hume thinks that both these two ideas are ‘perfectly distinct’, must concede that he is taking these two ideas to be sufficiently intelligible to be available for use in an informative description of a situation in which he would not face the philosophical difficulty he feels he does face.19 For Hume, ‘unintelligible’ means ‘not understandable’, ‘incomprehensible’. It does not mean ‘incoherent’, and so necessarily non-existent, as it standardly does today in philosophy. It means ‘not such that it has any empirically warrantable applicability to concrete reality’, hence ‘not clear in such a

In fact, ‘the same simple ideas [do] fall regularly into complex ones’, and one idea ‘naturally introduce[s]’ another (THN 1.1.4.2 / 11). This is what actually happens, and it cannot possibly happen, Hume says, unless there exists, as a matter of objective fact, a ‘bond of union’ – a ‘uniting principle’, ‘principles of union or cohesion’ – among our ideas (THN 1.1.4.1 / 10, 1.1.4.6 / 13). The ‘causes’ of this phenomenon are, he says, ‘mostly unknown, and must be resolv’d into original qualities of human nature, which I pretend not to explain’.22 But this is not to say that these principles of cohesion are not real. On the contrary, they are indeed real. All we can know of them are the observable regularities to which they give rise. Garrett states the general point robustly:

entirely loose and unconnected, chance alone wou’d join them; and ’tis impossible the same simple ideas should fall regularly into complex ones (as they commonly do) without some bond of union among them, some associating quality, by which one idea naturally introduces another.21

Hume is not forbidden by his empiricist principles from postulating the existence of unperceived deterministic mechanisms that would underlie the propensities of perceptions to appear in particular ways. He is forbidden by his principles only from trying to specify the nature of those mechanisms [in a way that goes] beyond what experience can warrant.23

188

9780826443595_Ch08_Final_txt_print.indd 188

2/9/2012 9:35:44 PM

HUME ON THE MIND So it is that when he is discussing causal necessity, Hume says that he is indeed, ready to allow, that there may be several qualities both in material and immaterial objects, with which we are utterly unacquainted; and if we please to call these power or efficacy, ’twill be of little consequence to the world. But when, instead of meaning these unknown qualities, we make the terms of power and efficacy signify something, of which we have a clear idea, and which is incompatible with those objects, to which we apply it [it is incompatible because it is just a feeling or impression], obscurity and error begin then to take place, and we are led astray by a false philosophy. (THN 1.3.14.27 / 168)24

8. GARRETT’S OBJECTION Many Hume commentators would reject Garrett’s claim about Hume in this passage. I accept it, for reasons given at the end of section 2 above, in spite of the fact that it suggests the following objection to the present account of Hume’s problem. Look, as you say, Hume holds that ‘the essence of the mind [is] unknown’ (THN Intro. 8 / xvii). He takes it for granted, in his philosophy considered as a whole, that there is in fact something more to the mind than just a series or bundle of perceptions. He never endorses the bundle theory of mind as the truth about the ultimate nature of the mind. This is all true. But this ‘something more’ is not a problem for him, contrary to what you suggest. It is not a problem for him because he can treat it in the same way in which he treats many other things – as something not further explicable by us, something ‘mostly unknown’ that ‘must be resolv’d into original qualities

of human nature, which I pretend not to explain’ (THN 1.1.4.6 / 13), something that is part of the unknown essence of the mind, something ‘wonderful and unintelligible’ (i.e. not understandable) by us, as he says of reason (THN 1.3.16.9 / 179), something ‘magical’, as he says of the Imagination (THN 1.1.7.15 / 24). This is a good objection, and it is raised explicitly, in one form, by Garrett, so I will call it ‘Garrett’s objection’.25 The principal difficulty for it can be put by saying that it seems to be an objection that must be put to Hume himself, because it is Hume himself who so plainly says that he has the problem that he does not have if Garrett’s objection is correct. When exactly do Hume’s hopes vanish? They vanish when he comes ‘to explain (P3) the principle of connexion, which binds . . . our particular perceptions . . . together’, ‘to explain (P4) the principles, that unite our successive perceptions in our thought or consciousness’ (THN App. 20–1 / 635–6). What does he mean by ‘explain the principle of connexion which binds’ or ‘explain the principles that unite’? What failure of explanation does he have in mind? He tells us. It is his failure to explain the principles of connection that make us ‘attribute . . . a real simplicity and identity’ to our perceptions. It is his failure, in other words, to explain the existence and operation of the principles of the Imagination – I will call the whole set of them the I-Principles, for short – that lead us to come to believe in a single continuing mind or self or subject. The problem, as he sees it, is that he cannot make use of the fact of the existence of the I-Principles in his philosophy without thereby appealing to – or rather, without being open to the charge that he thereby appeals to – something he cannot appeal to (O1), the idea of the mind as

189

9780826443595_Ch08_Final_txt_print.indd 189

2/9/2012 9:35:44 PM

HUME ON THE MIND ‘something simple and individual’, or (O2), some perceivable or experienceable ‘real connexion’ between perceptions.26 The lamented failure, then, is a failure to explain the existence of the I-Principles given the resources of the empirically warranted account of the mind. It is not a failure to explain how the I-Principles – in particular the Resemblance principle and the Cause and Effect principle – lead us to come to believe in a persisting mind or self.

9. COULD HUME’S PROBLEM BE THE ‘PROBLEM OF DETAIL’? Many have supposed that this is the lamented failure. Subtle philosophers have done so. So let us grant for the moment that the ‘when I come to explain . . .’ passage can be read in this way, at least when it is taken in isolation from the rest of the text. Hume, then, is despairing of his account of how the idea of a persisting self arises in us, on the grounds that the I-Principles (in particular the Resemblance and Cause and Effect principles of association) cannot really do the job. I will say that on this view Hume’s problem is the Problem of Detail. Is this a defensible interpretation (loss of all hope seems a strangely extravagant reaction to such a problem)? The way to find out is to look at what he thinks might solve his problem. But now we are back on the track already laid out in section 3. One thing that will do the trick is (the right to appeal to) the existence of (O1), ‘something simple and individual’ in which our perceptions inhere; another is (O2), some perceivable or observable ‘real connexion’ between perceptions,27 perceivable in such a way that the deployment of the idea of it in one’s philosophy when treating

of concrete reality is empirically warranted. Both (O1) and (O2) will provide Hume with the resources to explain the thing he has just said he cannot explain. This, however, rules out the view that his problem is the Problem of Detail. For neither (O1) nor (O2) can help the Resemblance and Cause and Effect principles of the association of ideas do their job in explaining an unwarranted belief in a persisting mind or self (for a doubt, see Strawson, The Evident Connection, p. 137). Again this is a somewhat backwards way to put the point. A better way to put it, perhaps, is to say that what destroys Hume’s hopes is his realization that he cannot meet the objection that he has in effect appealed to one of (O1) and (O2) in placing the I-Principles at the very centre of his theory of human nature, in making them the great engine of his philosophy. He has in effect appealed to one of (O1) and (O2) although he cannot appeal to either on his own terms. One could put the point by saying that the passage parses like this: ‘. . . when I proceed to explain [the principle of connexion, which binds them together, and makes us attribute to them a real simplicity and identity]; I am sensible, that my account is very defective . . .’. The noun-clause inside the square brackets denotes his problem, the phenomenon that is to be explained (=[P3/P4]). His problem is to explain the phenomenon that consists in the mind’s operating according to the I-principles, not to explain how the I-principles (in particular Resemblance and Causation), once in place, can generate the belief in ‘real simplicity and identity’. For, once again, the two ontological solutions to his problem that he offers in the next paragraph cannot be construed as a solution to that problem (the Problem of Detail). One could say that the word ‘explain’ is misread. It does not mean ‘expound’ or ‘spell

190

9780826443595_Ch08_Final_txt_print.indd 190

2/9/2012 9:35:44 PM

HUME ON THE MIND out’ – expound or spell out the details of how the principle of connection that makes us attribute simplicity and identity to our perceptions does its job. It means ‘account for the existence of’: account for the existence and operation of the I-principles given the resources of a strictly empiricist account of the mind. Hume’s problem is to give an account of the existence of the principle of connection that does what it does, given his commitment to the view that the bundle view is for philosophical purposes ‘the true idea of the human mind’. This cannot be done.28

10. ‘EXPLAIN’? – No. Hume can and does treat the phenomenon of the existence and operation of the I-Principles in the same way that he treats other things, as something not further explicable by us, something ‘mostly unknown’ that ‘must be resolv’d into original qualities of human nature, which I pretend not to explain’ (THN 1.1.4.6 / 13). The very fact that Hume uses the word ‘explain’ in the two crucial passages – ‘when I proceed to explain the principle of connexion, which binds them together’ and ‘when I come to explain the principles, that unite our successive perceptions in our thought or consciousness’ – proves that your interpretation cannot be right. For your interpretation requires us to suppose that Hume is lamenting his inability to explain something that he has repeatedly said he cannot and does not need to explain.

This is Garrett’s objection, somewhat extended. The best thing to do by way of reply, I think, is to start by considering points of agreement.

In operating in a way that is correctly described by the I-principles, the mind delivers all sorts of unity-and-connection experiences, which we may call UC experiences for short. It delivers persisting-physicalobject unity-and-connection experiences, it delivers necessary-causal-relation unityand-connection experiences, and it delivers persisting-individual-self unity-and-connection experiences. Is the existence of such experiences problematic for Hume? Not at all. He can fully explain the fact that we naturally believe in these sorts of unity and connection, even if our basic experience consists of nothing more than a series of distinct and fleeting perceptions, by appeal to the idea that the mind operates according to certain principles – the I-principles – that generate such (‘fiction’-involving) UC experiences. We cannot, however, explain the undoubtedly real phenomena to which we refer when we speak of the operation of the mind in accordance with the I-principles by reference to the operation of the mind in accordance with the I-principles, any more than we can use logic to prove the validity of logic. So the fact that the mind operates in accordance with the I-principles must be taken as a given (exactly as the conformity of physical phenomena to Newton’s law of gravity is taken as a given). Right. This is what Hume does. The causes of the mind’s operation in accordance with the I-principles are, he says, ‘mostly unknown, and must be resolv’d into original qualities of human nature, which I pretend not to explain’.29 He could hardly be more clear: ‘to explain the ultimate causes of our mental actions is impossible’ (THN 1.1.7.11 / 22). Much is unknown, then, and must remain so. So far Hume, Garrett and I fully agree. And Garrett and I also agree – contrary to a cloud of commentators – that Hume does in fact appeal to real connections throughout

191

9780826443595_Ch08_Final_txt_print.indd 191

2/9/2012 9:35:44 PM

HUME ON THE MIND Book I of the Treatise in appealing as he does to the I-Principles – to the ‘uniting principle’ or ‘bond of union’ that exists – the ‘uniting principles’ that exist – between our different perceptions.30 It is also plain that, pre-Appendix, Hume thinks that he can do this with impunity within his empiricist philosophy, because he can comfortably consign that in virtue of which the I-Principles exist to the ‘unknown essence of the mind’ in a thoroughly and indeed quintessentially Newtonian spirit.

11. REPLY TO GARRETT’S OBJECTION It is at this point that the agreement ends. For I think, as Garrett does not, that Hume’s hopes vanish when he sees that there is an objection from which this large confession of ignorance – this affirmation of ignorance – cannot protect him.31 The affirmation of ignorance sweeps up almost everything, but it leaves a hole. Hume’s position is vulnerable to the charge that if one relies on (P3/P4) – if one relies on the idea of the mind’s operation in accordance with the I-principles – then one is obliged to accept that one of the two maximally general positive metaphysical characterizations of the mind’s nature ([O1] or [O2]) must apply. But to accept this is to accept that one must allow the applicability of terms that are ‘unintelligible’ by Hume’s empiricist principles (P1) and (P2). This is how (P3/P4), (P1) and (P2), and (O1) and (O2) relate.32 Acknowledgement that one of maximally general (O1) or (O2) must be the case is compatible with vast ignorance of the nature of things, but Hume needs one of them, for he will not otherwise be able to ‘explain’ the I-Principles ([P3/P4]) in the following highly general sense: he will not be able to account

for how they exist at all. So he will not be able to make use of the idea that they exist. But they are central to his philosophy. Note that to give an explanation of something X in this sense, to give an account of things that makes room for the bare fact of X’s possibility, is not to attempt any further detailed explanation of X of the sort Hume thinks is impossible and is happy to leave as unknown. There is a moment when it dawns sharply on Hume that he has a problem. He realizes that the maximally general objection that one of (O1) and (O2), at least, is needed, and must in effect be allowed, given his account of the mind, can be most powerfully pressed against him. I suspect that it was the idea of others coming up with this objection that was most vivid for him as he wrote the Appendix. One thing he then wanted to do, most understandably, was to be the first to make the criticism (compare Wittgenstein’s assault on his earlier position). His best defence was to show complete candour and to be the first to describe the fork – the either-a-singlething-or-perceivable-real-connection fork – on which others would seek to spike him. Imagine how you yourself would feel, and what you might wish to do, if you discovered a serious difficulty in your just published and cherished theory. You would sit down and do something comparable to what Hume did when (probably hastily) he added the passage on personal identity to the Appendix. One could put the point by saying that the existence and operation of the I-principles mean that some metaphysical description of the mind that Hume cannot avail himself of is knowably applicable to the mind. He cannot invoke the mind’s ‘unknown essence’, treating this as a kind of explanation-sink that can absorb the whole difficulty, for – this is the direct reply to Garrett’s objection – his

192

9780826443595_Ch08_Final_txt_print.indd 192

2/9/2012 9:35:44 PM

HUME ON THE MIND opponents can happily grant that of course much must remain unknown, while continuing to insist that Hume has, in appealing to the I-Principles, invoked something – some sort of genuine metaphysical connection and continuity among the perceptions of the mind – whose existence he can make sense of only on one of two specific conditions, neither of which is available to him. It is hardly impressive (it is hopeless) for him, faced with such an objection, to answer again that much is ‘unknown’, ‘magical’, ‘unintelligible’, ‘wonderful’ and ‘inexplicable’. ‘Yes, yes,’ his objectors reply in turn, ‘we agree. The point we wish to make is much more general (it is, in twentieth-century parlance, a “logical” point). In relying on the I-Principles as you do you take a metaphysical step you cannot take, given that you want to give an empiricist account of the mind as well as everything else. You incur a certain general metaphysical debt you cannot repay on your own empiricist principles. You cannot rely on the I-Principles as you do and simply refer everything else to the unknown essence of the mind, for you cannot stop someone replying that your reliance on the I-Principles entails that there is at least one thing that can be known about the essence of the mind and that you cannot allow to be known. The thing in question is in fact an either-or thing ([O1] or [O2]), but that does not help. You cannot allow this either-or thing to be known, because it is not possible to specify what it is without employing terms whose employment you cannot allow, given your brand of empiricism, when it comes to making knowledge claims about the nature of concrete reality.’ ‘Specifically, and once again, your reliance on the I-Principles entails that the following high-level, either-or description of the essence of the mind – “persisting individual single thing or really connected plurality of

things” – can be known to apply. You cannot make room for this because you cannot allow any empirical meaning or (therefore) concrete applicability to any idea of anything whose description entails that it lasts longer than a single fleeting perception. A fortiori you cannot admit that any such idea has an indispensable employment in your philosophy, or that your philosophy presupposes that such an idea has valid application. But it does. Your philosophy entails – we are hammering the point – that we can know at least one thing more about the essence of the mind than you say we do or can: we can know something that we cannot and must not claim to know on your empiricist principles. How else can it possibly be the case that perceptions come clumped in interacting groups as they do?’ To this Hume thinks, quite rightly, I believe, that he has no effective reply. He cannot say what he actually believes, given the dialectical context of his discussion of personal identity. He cannot say that the brain supplies all the needed real continuity. And even if he did, this would not diminish his need to acknowledge real connection, for the brain is certainly not a simple substance (which is, after all, a property reserved to individual atoms and immaterial souls).

12. A FINAL RESPONSE – You are seriously underestimating Hume’s resources. He is ‘not forbidden by his empiricist principles from postulating the existence of unperceived deterministic mechanisms that would underlie the propensities of perceptions to appear in particular ways. He is forbidden by his principles only from trying to specify the nature of those mechanisms [in a way that goes] beyond what experience

193

9780826443595_Ch08_Final_txt_print.indd 193

2/9/2012 9:35:44 PM

HUME ON THE MIND can warrant’. But he does not try to do this, in the case of the mind, nor does he think he needs to. He is, again, happy to say that what you call the ‘I-principles’ are unintelligible, inexplicable and wonderful. He has, therefore, no problem of the sort you describe. This is another version of Garrett’s objection, mostly in his own words.33 I think I have answered it. Hume does not think he can plausibly reject the objection that he is committed to something like (O1) or (O2), caught in a fork according to which one at least of (O1) or (O2) is correct (he is caught because it is a maximally general and exhaustive fork). (O1) and (O2) are very general, but when we consider the mechanisms to which Hume can legitimately appeal, while holding them to be unknown, we see that (O1) and (O2) already ‘specify the nature [or ground] of those mechanisms’ in a metaphysical way that goes ‘beyond what experience can warrant’. In conclusion, let me repeat the earlier suggestion that Garrett’s objection has to be put to Hume himself, because it is Hume himself who thinks he has a problem that could be entirely solved if he were allowed to make use of the idea of a simple individual substance, or the idea of (empirically observable) real connections. This is the fundamental fixed point, when it comes to the interpretation of the Appendix. It is Hume himself who judges (sees) that he is in effect committed, in his philosophy, to the allowability of at least one of two very high-level metaphysical descriptions of the nature of the mind that can have no empirical warrant and are therefore officially excluded from any role in his philosophy. It is Hume himself who thinks that his empiricism allows him to ignore (delegate to the unknown, be agnostic about) all questions about the ultimate causes or sources

of the patterns in our experiences that lead us to come to believe in physical objects and causal necessity, but does not allow him to do this when it comes to the mind itself. It is Hume himself who thinks he has a problem he cannot solve even after he has stressed the unintelligibility and inexplicability (and ‘wonderfulness’ and ‘magicality’) (THN 1.3.16.9 / 179, 1.1.7.15 / 24) of the workings of the mind – the mind whose principles of working are the great and indispensable engine of his whole empiricist programme – and who (again) thinks that he could solve the problem immediately if the principles of his philosophy allowed him to deploy the notion of a simple and individual substance, or to make empirically warranted use of the notion of real (non-‘fictional’, non-Imagination-generated) connections. The burden on those who favour what I am calling Garrett’s objection is to explain why Hume feels he has a problem he could solve if he could appeal to a persisting individual substance or make use of an empirically warranted notion of real connection. It is Hume himself who believes himself to be in a Zugzwang – a position where he would like to be able to make no move but feels he is obliged to make one (or admit that he has in effect already made one). Old interpretative impulses may resurge: ‘For Hume, the phenomenon of conformity to the I-principles is brute regularity; there is therefore no need or possibility of any further explanation of any sort, however general, in his scheme.’ But it is far too late in the day for such a view of Hume, and there are two more particular replies. First, his problem stems from the fact that he has ‘loosen’d all our particular perceptions’ (THN App. 20 / 635); but this loosening would not cause a problem if he took it that the phenomenon of conformity to the I-principles were just a matter

194

9780826443595_Ch08_Final_txt_print.indd 194

2/9/2012 9:35:44 PM

HUME ON THE MIND of brute regularity. Secondly, a reply already made. You have to contrapose: it is Hume himself who insists that the phenomenon of conformity to the I-principles does need some further explanation or grounding, however general, and who tells us that two things that are completely unavailable to him would do the trick: inherence in a single substance or real – non-regularity-theory – connections. It is not as if he wants to say any such thing, appealing to notions whose use in philosophy he has ruled out as ‘unintelligible’.34 It is just that he believes (sees) that the objection that he must admit some such thing is correct and unanswerable. When he moved on from (3) his empiricist account of the content of the idea of causation in 1.3.14 / 155–72 of the Treatise, and (5) his empiricist account of the content of the idea of physical objects in 1.4.2 / 187–218, and took (1) the empirically warranted idea of the mind itself as his subject in 1.4.6 / 251–63, his general ‘reductive’ empiricist account of the origin of our belief in the objective continuities, persistences and connections that we take ourselves to encounter in experience was running beautifully. It was watertight on its own terms, and it must have seemed that it could not fail to deal also with the apparent or experienced continuity of the mind. And in a sense it did, and smoothly too: it gave at least as good an account of the origin of (2) our idea of ourselves as enduring selves or subjects as it did of the origin of (4) our ideas of causal power and (6) our ideas of physical objects (which is not to say that it was in fact empirically psychologically correct). But it relied on something more than (1) could supply. The whole system broke down when it came to (1), the empirically warranted idea of the mind. It is Hume himself – one more time – who believes that his account of the mind is ‘very defective’, indeed hopeless, and it is Hume himself who believes that

his problem would be immediately solved by one of two metaphysical provisions that his empiricist philosophy rules out.35

NOTES 1

2

In Book I of the Treatise Hume takes ‘person’ to have a merely mental reference, and uses it interchangeably with ‘mind’ and ‘self’ (and sometimes ‘soul’). See, for example, N. Pike, ‘Hume’s Bundle Theory of the Self: A Limited Defense’, American Philosophical Quarterly 4 (1967), pp. 159–65: ‘when Hume uses the term . . . “person”, he generally means to be referring only to the mind’ (p. 161), at least in Book I of the Treatise. See also THN 1.4.6.2 / 251, 1.4.6.5 / 253, 1.4.6.20 / 262, ‘self or person’; 1.4.6.17 / 260, ‘mind or thinking person’. Most commentators have thought his problem lies in his account of [2]. See J. Ellis, ‘The Contents of Hume’s Appendix and the Source of His Despair’, Hume Studies 32 (2006), pp. 195– 231, for an interesting recent defence of this view. See also H. Noonan, this vol., pp. 178–9. For the opposing view (other than this paper), see, for example, B. Stroud, Hume (London: Routledge & Kegan Paul, 1977) and, more recently, D. Garrett, ‘Rethinking Hume’s Second Thoughts about Personal Identity’, in J. Bridges, N. Kolodny and W. Wong (eds), The Possibility of Philosophical Understanding: Essays for Barry Stroud (New York: Oxford University Press, 2011); G. Strawson, The Evident Connexion: Hume on Personal Identity (Oxford: Oxford University Press, 2011), sects. 3.4, 3.10. Hume later comes to think that Kames gives a better account of the origin of [2] than he does. Reading a draft of Kames’s Essays in 1746, Hume writes to Kames that ‘I likt exceedingly your Method of explaining personal Identity as more satisfactory than any thing that had ever occurr’d to me’ (NLH 20, 8). I suspect that Hume here means Kames’s account of the origin of our idea of or belief in a persisting self – ‘man . . . has an original feeling, or consciousness of himself, and of his existence, which for the most part accompanies every one of his impressions

195

9780826443595_Ch08_Final_txt_print.indd 195

2/9/2012 9:35:45 PM

HUME ON THE MIND

3

4

5

6

7

8

and ideas, and every action of his mind and body’ (Lord Kames, ‘Of the Idea of Self and Personal Identity’ in Essays on the Principles of Morality and Natural Religion (Edinburgh: Fleming, 1751), pp. 231–2) – if only because Kames’s further remarks (e.g. ‘this consciousness or perception of self is, at the same time, of the liveliest kind. Self-preservation is everyone’s peculiar duty; and the vivacity of this perception, is necessary to make us attentive to our own interest’ (ibid., p. 232)) are very close to Hume’s own published views in Books 2 and 3 of the Treatise (see, for example, THN 2.1.11.4 / 317; 2.1.11.8 / 320; 2.2.2.15–16 / 339–40; 2.2.4.7 / 354; 2.3.7.1 / 427). Hume holds, of course, that only an empiricist philosophy deals in clear and distinct ideas. He uses ‘clear and distinct’ a couple of times in the Treatise when discussing geometrical concepts (THN 1.1.7.6 / 19, 1.2.4.11 / 43). He uses ‘clear and distinct . . . idea’ and ‘clear, distinct idea’ in the Enquiry (see EHU 12.20 / 157, 12.28 / 164), also ‘clearly and distinctly’ (EHU 4.18 / 35); otherwise he uses ‘clear’ and ‘distinct’ separately. He also uses ‘clear and precise’ (Abs. 7 / 648, THN 1.3.1.7 / 52). All mental occurrences are perceptions, in Hume’s terminology – thoughts, sensations, emotions, ideas, and so on – and they are all (by definition) conscious. The word that now corresponds most closely to Hume’s word ‘perceptions’, in this use, is ‘experiences’. Here the ‘causes and effects’ are particular perceptions. THN 1.4.6.4 / 252, 1.4.6.4 / 253, 1.4.6.18 / 260, 1.4.6.20 / 262, App. 15 / 634, App. 20 / 635, 1.4.2.39 / 207, 1.4.2.40 / 207, 1.4.7.3 / 265, Abs. 28 / 658. Six of these eleven quotations are from passages where Hume is discussing something else, or summarizing the view and stating it in a particularly compressed form. See also THN 2.1.2.2 / 277. On this see in particular E. Craig, The Mind of God and the Works of Man (Cambridge: Cambridge University Press, 1987), pp. 111–20. As Garrett remarks, the question of what Hume’s problem was – the question of what Hume thought his problem was – ‘has received what is surely a far greater number

of distinct answers – well over two dozen, even by a conservative count – than has any other interpretive question about Hume’s philosophical writings’ (‘Rethinking Hume’s Second Thoughts’, p. 16). I do not think this would have happened if the discussion had started out from THN 1.1.4 / 10. 9 In speaking of attraction in the natural world he is referring to gravity. 10 The reference to ‘simple’ in ‘the principles of union or cohesion among our simple ideas’ could be dropped. 11 There is a footnote reference letter (‘a’) attached to the word ‘when’ in Hume’s text. The note refers the reader to ‘Vol. I. Page 452. This falls wholly on p. 260 in the Selby-Bigge edition, beginning with ‘. . . if disjoin’d by the greatest difference’ (THN 1.4.6.16) and ending with ‘. . . amidst all its variations’ (THN 1.4.6.18). 12 Note that the problem is not just that he cannot give an account of the fact that ‘the same simple ideas should fall regularly into complex ones’. That is simply the first point he considers in THN 1.1.4 / 10–13, ‘Of the connexion or association of ideas’. He also needs to give an account of the operation of the other associations of ideas that the ‘uniting principle’ accounts for, the associations of ideas based on resemblance, contiguity and causation, which he sets out in the rest of the section. (I think Hume may have turned to this section to guide his words when sketching his difficulty in the Appendix.) 13 As Garrett says, ‘by “real connexion” used as a technical term, Hume means (at least) a connection between two objects that is more than simply an associative relation in the imagination’ (Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997), p. 181). Hume’s principal example of a ‘real connexion’ is causal necessity realistically and naively figured as something that exists quite independently of any activity of the fictiongenerating ‘imagination’. 14 Putting aside the fact that Hume’s psychological account is wholly inadequate, empirically speaking. See, for example, Hermer, L. and Spelke, E., ‘Modularity and Development: The Case of Spatial Reorientation’, Cognition

196

9780826443595_Ch08_Final_txt_print.indd 196

2/9/2012 9:35:45 PM

HUME ON THE MIND

15

16

17

61 (1996), pp. 195–232; Carey, S., The Origin of Concepts (Oxford: Oxford University Press, 2009), chap. 3. For independent proof that Hume takes there to be more to the mind than perceptions, see, for example, G. Strawson, The Secret Connexion: Causation, Realism, and David Hume (New York: Oxford University Press, 1989), p. 130; Strawson, The Evident Connexion, sect. 2.6. (Briefly: call a simple impression of A an A-impression, and a simple idea of A an A-idea. According to Hume, an A-idea can arise in my mind only if I have already had an A-impression. What happens is that ‘there is a copy taken by the mind, which remains after the impression ceases; and this we call an idea’ (THN 1.1.2.1 / 8). But if the mind is just a bundle of distinct experiences with no hidden content (their contents are ‘perfectly known’), then there is no possible way in which this can happen. For where does the A-idea ‘remain . . . after the impression ceases’ – given that I then go on to experience or think about, B, C, and many other things, and have no conscious thought of A?) See also EHU 5.14 / 50: ‘nature has established connexions among particular ideas. . . . These principles of connexion or association we have reduced to three, namely, Resemblance, Contiguity and Causation; which are the only bonds that unite our thoughts together . . .’. This proposed solution comfortably satisfies four of the five criteria for a successful solution that Garrett lists in ‘Once More Into the Labyrinth: Kail’s Realist Explanation of Hume’s Second Thoughts about Personal Identity’, Hume Studies 36(1) (2010), pp. 77–87: the ‘Crisis Criterion’, the ‘Origin Criterion’, the ‘Solution Criterion’, and the ‘Scope Criterion’. It questions whether the fifth criterion – the ‘Difficulty Criterion’ – is correct, by proposing that Hume did not have any difficulty in stating his problem – although he could certainly have been clearer. The explanation of why Hume has been found obscure lies in the preconceived ideas about Hume that readers have brought to the Appendix. (The clause that has been most damagingly misread is ‘when I proceed to explain the principle of connexion, which binds them together, and makes us

18

19

20

21

22

23

24

25

attribute to them a real simplicity and identity’. See sects. 8 and 9 below.) ‘We have a distinct idea of an object, that remains invariable and uninterrupted thro’ a suppos’d variation of time; and this idea we call that of identity or sameness. We have also a distinct idea of several different objects existing in succession, and connected together by a close relation; . . . these two ideas of identity, and a succession of related objects [are] in themselves perfectly distinct’ (THN 1.4.6.6 / 253). Compare the move he makes in THN 1.4.5 / 232–51, discussed in Strawson, The Evident Connexion, p.50. Craig has an incomparable discussion of this issue; see in particular Craig, The Mind of God and the Works of Man, pp. 123–30, a decisive antidote to Millican’s unfortunate recent attempts to re-equate Hume’s use of ‘unintelligible’ with ‘incoherent’ (see, for example, P. Millican, ‘Hume, Causal Realism, and Causal Science’, Mind 118 (2009), pp. 647–712, pp. 647–8). See also Strawson, The Secret Connexion, pp. 49–58 and Strawson, The Evident Connexion, sect. 2.4. THN 1.1.4.1 / 10; emphasis added. See Strawson, The Evident Connexion, pp. 140–1 for a discussion of the point that these mental connections involve only a ‘gentle force’, and are not exceptionless. THN 1.1.4.6 / 13. He follows Newton, who states the law of gravitational attraction while adding that ‘the cause of Gravity . . . I do not pretend to know’ (Principia, trans. A. Motte and F. Cajori (Berkeley: University of California Press, 1934 [1687]), 3.240). Garrett, Cognition and Commitment, p. 171; emphasis added. Notice the relative mildness of ‘obscurity and error then begin’. Compare THN 2.3.2.4 / 409–10. See Garrett, Cognition and Commitment, p. 171. It’s raised as an objection to Stroud, Hume and T. Beauchamp, ‘Self Inconsistency or Mere Self Perplexity?’, Hume Studies 5 (1979), pp. 37–44. Garrett has changed his view about Hume’s problem in the Appendix since he published Cognition and Commitment in 1997 (see Garrett, ‘Once More into the Labyrinth’), and he and I are in agreement in one central respect.

197

9780826443595_Ch08_Final_txt_print.indd 197

2/9/2012 9:35:45 PM

HUME ON THE MIND 26

27

28

29

30

A full account of the I-Principles – the principles according to which the Imagination operates – must go beyond the three principles of the association of ideas (Resemblance, Contiguity and Cause and Effect) and add the fundamental principle according to which the Imagination is unfailingly led to posit or ‘feign’ objective continuities (persisting objects, a persisting individual mind, and true causal continuities) on the basis of exposure to certain sorts of sets or series of ideas. See Strawson, The Evident Connexion, sect. 3.4. Some connection which is not just an I-principles-generated connection in the Imagination, and is, therefore, essentially more than the relation of cause and effect so far as we have any empirically contentful notion of it. (As remarked in n. 13 above, Hume’s prime example of ‘real connexion’ is causal necessity thought of as something that obtains quite independently of any action of the imagination.) The clause ‘which binds them together’ can be read in two ways. On one reading it is about what the I-principles do: they lead us to put or bind experiences together in such a way as to take them to be parts or rather features of a single continuing object. According to the other reading it is about the phenomenon of our experiences being actually bound together – united, connected – in being governed by the I-principles. According to this second reading it is only the clause ‘which makes us attribute to them’ that is about what the I-principles do. Either way, the point remains: that the thing that Hume has to explain is the existence of ‘the principle of connexion’, not how it does what it does. THN 1.1.4.6 / 13. Hume would have loved modern neuroscience – although not as much as Descartes. THN 1.1.4.1 / 11, 1.4.6.16 / 260. Hume also mentions the ‘uniting principle among our internal perceptions’ at the heart of his

31

32

33

34

35

principal discussion of causation (THN 1.3.14.29 / 169). It is a matter of unintelligible real connection in just the same way as the uniting principle ‘among external objects’. For Garrett’s own account of Hume’s problem in the Appendix, in terms of ‘placeless perceptions’, see Garrett, ‘Once More into the Labyrinth’. All six occur in the space of 87 words (of which they make up 48): ‘. . . all my hopes vanish, when I come to explain ([P3/P4]) the principles, that unite our successive perceptions in our thought or consciousness. I cannot discover any theory, which gives me satisfaction on this head. In short there are two principles, which I cannot render consistent; nor is it in my power to renounce either of them, viz. [P1] that all our distinct perceptions are distinct existences, and [P2] that the mind never perceives any real connexion among distinct existences. Did [O1] our perceptions either inhere in something simple and individual, or [O2] did the mind perceive some real connexion among them, there wou’d be no difficulty in the case.’ (THN App. 20–1 / 635–6; emphasis added) Garrett, Cognition and Commitment, p. 171. Of course Garrett’s text long predates, and is not a response to, the present one. Recall again that he uses them constantly in a way that presupposes that they do have content and are to that extent intelligible, and does not mean what present-day philosophers mean by ‘unintelligible’. This paper descends from ‘Hume on Himself’, a paper given at the Hume Society conference in Cork in 1999 and published ‘too precipately’ in 2001 (in D. Egonsson, J. Josefsson, B. Petersson and T. Rønnow-Rasmussen (eds), Essays in Practical Philosophy: From Action to Values (Aldershot: Ashgate Press, 2001), pp. 69–94). I am grateful to Don Garrett for comments on a later (2003) version, and to Stephen Buckle for his comments on this one.

198

9780826443595_Ch08_Final_txt_print.indd 198

2/9/2012 9:35:45 PM

9 ACTION, REASON AND THE PASSIONS Constantine Sandis

1. ACTION AND ITS CAUSES Actions, for Hume, are external objects in the sense of being things which we can observe through the senses. Our knowledge of them is therefore not a priori but empirical, mediated as it is through perceptual impressions. Accordingly, Hume believes that purported explanations of action, be they singular or general, are to be tested through experience, either directly or through testimony, for ‘we can give no reason for our most general and most refin’d principles, beside our experience of their reality’ (THN Intro.10 / xviii). This does not entail that the reasons for which we act are themselves external, observable, objects.1 Rather, their existence is to be inferred from behaviour, the power of any given argument from analogy hanging on the proper degree and nature of philosophical scepticism about causal reasoning. Reasons why people did or believed certain things figure on virtually every single page of all six volumes of The History of England. Hume also mentions such reasons in his philosophical works, be it explicitly (e.g. THN 2.2.5.4 / 358, 3.2.1.9 / 379) or implicitly (e.g. THN 1.3.4.2 / 83). He also describes reasons we have for acting (e.g. THN 1.3.9.13 / 133), making no ontological distinction between the latter and the former

kinds of reasons. Nonetheless, his view of human nature is highly alert to our tendency to over-rationalize actions, beliefs and passions that are typically a matter of habit, custom or sentiment (e.g. THN 2.2.3.9 / 351; 1.3.7.6 / 97). Indeed, Hume’s naturalist concept of what contemporary philosophers call normative reasons is proto-Wittgensteinian in so far as it is to be explained by reference to human propensities and practices, for example, expectation and induction, rather than the other way round (THN 1.3.6.3 / 88).2 According to Hume’s account, we may acquire knowledge of another person’s reasons or motives through a combination of inductive and analogical reasoning relating their behaviour to past instances: [I]n judging the actions of men we must proceed upon the same maxims, as when we reason concerning external objects. When any phænomena are constantly and invariably conjoin’d together, they acquire such a connexion in the imagination, that it passes from one to the other, without any doubt or hesitation. But below this there are many inferior degrees of evidence and probability, nor does one single contrariety of experiment entirely destroy all our reasoning. The mind balances the contrary experiments, and deducting the inferior from

199

9780826443595_Ch09_Final_txt_print.indd 199

2/9/2012 9:35:53 PM

ACTION, REASON AND THE PASSIONS Mankind are so much the same, in all times and places, that history informs us of nothing new or strange in this particular. Its chief use is only to discover the constant and universal principles of human nature, by showing men in all varieties of circumstances and situations, and furnishing us with materials, from which we may form our observations, and become acquainted with the regular springs of human action and behaviour. (EHU 8.7 / 83)

the superior, proceeds with that degree of assurance or evidence, which remains. Even when these contrary experiments are entirely equal, we remove not the notion of causes and necessity; but supposing that the usual contrariety proceeds from the operation of contrary and conceal’d causes, we conclude, that the chance or indifference lies only in our judgement on account of our imperfect knowledge, not in the things themselves, which are in every case equally necessary, tho’ to appearance not equally constant or certain. No union can be more constant and certain; than that of some actions with some motives and characters; and if in other cases the union is uncertain, ’tis no more than what happens in the operations of body, nor can we conclude anything from the one irregularity, which will not follow equally from the other (THN 2.3.1.12 / 403–4).

The work of Hume as a historian reveals the motivating influence of character. He embraces a moderately stoic virtue epistemology according to which the historian is in the emotionally privileged position to evaluate past actions correctly. This is to be achieved through an approximation of the golden mean between involved empathy and disinterested detachment:

The prediction and explanation of action thereby forms part of the science of human nature that Hume seeks to establish. Actions are no different from other events in being susceptible to scientific laws.3 As with natural science, explanation in social science is inductive not deductive: it is a matter of empirically informed conjectures. If there is to be any such thing as a logic of history, then, it is to be an inductive logic, the limitations of which Hume famously exposed. These conjectures may be based on patterns of reasoning as well as patterns of non-rational connections. What degree of certainty any given pattern entitles us to assume depends on whether one emphasizes Hume’s positive account of causal reasoning over his scepticism about causal reasoning, or vice versa. Either way, conjectures are to be confirmed or refuted through ‘cautious observation of human life’ (THN Intro. 10 / xix), the most systematic form of which is historiography:

When a man of business enters into life and action, he is more apt to consider the characters of men, as they have relation to his interest, than as they stand in themselves; and has his judgement warped on every occasion by the violence of his passion. When a philosopher contemplates characters and manners in his closet, the general abstract view of these objects leaves the mind so cold and unmoved, that the sentiments of nature have no room to play, and he scarce feels the difference between vice and virtue. History keeps in just medium betwixt these extremes, and places the objects in their true point of view. The writers of history, as well as the readers, are sufficiently interested in the characters and events, to have a lively sentiment of blame or praise; and, at the same time, have no particular interest or concern to pervert their judgement (E 568).4

200

9780826443595_Ch09_Final_txt_print.indd 200

2/9/2012 9:35:53 PM

ACTION, REASON AND THE PASSIONS For this reason Hume begins The History of England by noting that he will not concern himself much with either distant or recent history, for the former is too far removed from our concerns to be of any interest, and the latter too close for us to keep an impartial distance. His History thus begins with the Britons and ends with the last Stuarts (almost a century before the time in which it was written). Hume accordingly ends his essay ‘Of the Study of History’ with a quotation from Lucretius’ De Rerum Natura, translating roughly as ‘only then are the words of truth drawn up from the heart’ (E 568). In Lucretius, the context points to times of danger, peril or adversity whereas Hume is referring to that ‘just medium’ between lively sentiment and personal disinterest required for impartial judgement, a lesson more properly learned from Hume’s Stoic (‘the man of action and virtue’, E 146n1) than the Epicurean (‘the man of elegance and pleasure’, E 138n1). In stark contrast to the thought of Hume’s philosophical contemporaries, Hellenistic philosophy is not so abstract as to risk not feeling the sentiments of vice and virtue.5 Hume believed that the correct approach to human action is that of evaluating the character which the action reveals, it being a blatant falsehood that ‘all characters and actions [are] alike entitled to the affection and regard of everyone’ (EPM 1.2 / 169–70). This interest in character is temporarily forgotten in his discussion ‘of personal identity’,6 but the important role it plays in Books II and III of the Treatise, the second Enquiry, and Hume’s Essays and History suggest that it would be myopic for any account of Hume on the soul and the self to ignore the question of character.7 Hume would have agreed, for instance, with David Knowles’s pronouncement that ‘a life is not a bundle of acts; it is

a stream or a landscape; it is the manifestation of a single mind and personality that may grow more deformed or more beautiful to the end’.8, 9 Pace Hume, numerous writers on history including Croce, Knowles and Carr have claimed that it is not the business of the historian to pass moral judgements on individuals, but only to explain why they acted as they did.10 For Hume, however, a correct explanation of action will appeal to the agents’ motives, the discernment of which is necessarily evaluative: if you misjudge character you fail to explain action, for the appreciation of character and the understanding of action are inextricably tied together. A proper evaluation of character, Hume thinks, can determine the likelihood that any given event was the result of situation or temper. History tells us much about human nature but, conversely, the principles of human nature can, among other causal principles, help us to interpret history better. The History of England thereby aims ‘to provide an account of English history based on empirically plausible assumptions’,11 as opposed to those histories which fruitlessly invoke miracles, prophecy and revealed religion. Hume allows that actions may accord with more than one motive, just as Donald Davidson would later claim that they may accord with one or more reasons that the agent might have for action. According to Davidson, the criterion for determining which of the numerous reasons an agent might have for acting is the one he or she actually acted upon is causal (in a way which has proved to be highly problematic).12 By contrast Hume maintains, more pragmatically, that the correct method for attributing motives to any given individual is to ask which one(s) would reveal him as acting characteristically, a fact

201

9780826443595_Ch09_Final_txt_print.indd 201

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS to be determined on the purely empirical grounds of past regularity: [A]s the union betwixt motives and actions has the same constancy, as that in any natural operations, so its influence on the understanding is also the same, in determining us to infer the existence of one from that of another. If this shall appear, there is no known circumstance, that enters into the connexions and productions of the actions of matter, that is not to be found in all the operations of the mind; and consequently we cannot, without a manifest absurdity, attribute necessity to the one, and refuse it to the other . . . (THN 2.3.1.14 / 404) a spectator can commonly infer our actions from our motives and character; and even where he cannot, he concludes in general, that he might, were he perfectly acquainted with every circumstance of our situation and temper . . . (THN 2.3.2.2 / 408–9) in judging the actions of men we must proceed upon the same maxims, as when we reason concerning external objects . . . (THN 2.3.1.12 / 403) No union can be more constant and certain; than that of some actions with some motives and characters. (THN 2.3.1.12 / 404) The most irregular and unexpected resolutions of men may frequently be accounted for by those who know every particular circumstance of their character and situation. (EHU 8.15 / 88; cf. E 16) What neither reason nor human nature can explain is thereby attributed to character, which divides human beings into sorts.13 Christine Korsgaard has objected that the suggestion that agent-causation may be achieved ‘when the person’s character serves as a kind of filter in the causal

chain, making the outcome turn one way rather than another’ seems to ‘lose track’ of the ‘fact’ that nothing counts as an ‘action’ unless a person ‘is the cause of an intentional movement, or something of that sort’.14 It is true that as a philosopher Hume says nothing of agents causally determining intentional movements. As a historian, however, he takes it for granted that a strong character may determine the course of history, and he may well have been open to the Sellars-Davidson thesis that an action may be caused without its agent being caused to perform it.15 None of this leaves Hume oblivious to competing non-psychological causes of human action, as made clear in the following remark on political life and human nature: ‘So great is the force of laws, and of particular forms of government, and so little dependence have they on the humours and tempers of men, that consequences almost as general and certain may sometimes be deduced from them, as any which the mathematical sciences afford us’ (E 16). The case of law and government renders political events as close as human behaviour can come to naturally approximate events observed in controlled experiments. However, Hume’s deterministic science of behaviour can only be understood in the light of his understanding of causation and necessary connexion. This is not the place to try and settle ongoing interpretational disputes relating to the extent, if any, to which he espouses a ‘regularity’ theory of causation,16 but we might nonetheless note just how weak his definitions of ‘cause’ and ‘necessity’ actually are: ‘I define necessity in two ways, conformable to the two definitions of cause, of which it makes an essential part. I place it either in the constant union and conjunction of like objects, or in the inference of the mind from one to the

202

9780826443595_Ch09_Final_txt_print.indd 202

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS other’ (THN 2.3.2.4 / 409). Hume’s rejection of the doctrine that we possess a liberty of indifference (which he thinks of as the illusion that one’s actions are not causally necessitated by one’s motives) is thus more or less tantamount to a trivial truth. He writes: After we have perform’d any action; tho’ we confess we were influenc’d by particular views and motives; ’tis difficult for us to perswade ourselves we were govern’d by necessity, and that ’twas utterly impossible for us to have acted otherwise; the idea of necessity seeming to imply something of force, and violence, and constraint, of which we are not sensible . . . We may imagine we feel liberty within ourselves; but a spectator can commonly infer our actions from our motives and character; and even when he cannot, he concludes in general, that he might, were he perfectly acquainted with every circumstance of our situation and temper, and the most secret springs of our complexion and disposition. Now this is the very essence of necessity, according to the foregoing doctrine (THN 2.3.2.1–2 / 407–9).17 Human behaviour is as much the product of an unobservable causal necessity as any other natural event. The only difference between them is epistemic: our knowledge of the principles of human nature that bind motion to action is less precise than that of the ‘universally allowed’ deterministic laws which bind physical force to motion. This is partly due to the fact that the former laws are considerably more complicated, but it is equally a result of the fact that it is all but impossible to perform extensive controlled experiments involving human action (though Hume would have certainly been interested in the work of Benjamin Libet).

Be all this as it may, our imperfect psychophysical knowledge is nonetheless sufficient to enable us to predict individual and social behaviour in an indefinite number of situations. None of this prevents Hume from pursuing his ‘reconciling project’ of demonstrating that necessity (as he has defined it) is compatible with free will, which he equates to the liberty of spontaneity to do as one desires. Far from being an obstacle to moral responsibility, the necessity which binds character to action is required for its existence, at least given his account of the virtues according to which the viciousness or virtue of any given act arises from ‘some cause in the character and disposition of the person who performed them’ (EHU 8.29 / 98). Another corollary of Hume’s position is that freedom increases in proportion to madness: ’Tis commonly allow’d that mad-men have no liberty. But were we to judge by their actions, these have less regularity and constancy than the actions of wise-men and consequently are further remov’d from necessity. Our way of thinking in this particular is, therefore, absolutely inconsistent; but it is a natural consequence of these confus’d ideas and undefin’d terms, which we so commonly make use of in our reasonings, especially on the present subject. (THN 2.3.1.13 / 404) Given Hume’s definitions, the claim that free will and morality are compatible with causal necessity is unobjectionable. Hume asserts that ‘if anyone alters the definitions, I cannot pretend to argue with him, till I know the meaning he assigns to these terms’ (THN 2.3.1.18 / 407). Whether it is he or his opponents who are playing with words is, of course, another matter.

203

9780826443595_Ch09_Final_txt_print.indd 203

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS I have prov’d, that reason is perfectly inert, and can never either prevent or produce any action or affection (THN 3.1.1.8 / 458).

2. MOTIVATION, REASON AND BELIEF Hume famously claims that ‘[r]eason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them’ (THN 2.2.3.4 / 414). This remark, in tandem with his ‘influence argument’ to the conclusion that the rules of morality ‘are not the conclusion of our reason’ (THN 3.1.1.7 / 457), has spawned a hideous number of theses in moral psychology as diverse (and incompatible) as error theory, quasi-realism, expressivism, emotivism, prescriptivism, projectivism, non-cognitivism, reasons internalism, instrumentalism, hypotheticalism, contextualism, scepticism, egoism, relativism, subjectivism, motivation internalism, sentimentalism and the Humean theory of motivation. In what follows I focus on the last of these, only touching upon the rest (which are primarily concerned with issues relating to what has come to be called ‘the nature of moral judgement’)18 as and when they relate to everyday motivation. Hume repeatedly emphasizes the limitations of reason as a motivating power:19 Reason alone can never be a motive to any action of the will (THN 2.3.3.1 / 413). Abstract or demonstrative reasoning, therefore, never influences any of our actions, but only as it directs our judgement concerning causes and effects (THN 2.3.3.2 / 414). [I]mpulse arises not from reason but is only directed by it (THN 2.3.3.3 / 414). [R]eason alone can never produce any action, or give rise to volition . . . the same faculty is as incapable of preventing volition (THN 2.3.3.4 / 414–5). [R]eason has no influence on our passions (THN 3.1.1.7 / 457).

[R]eason can never immediately prevent or produce any action by contradicting or approving it . . . Reason is wholly inactive (THN 3.1.1.10 / 458).20 Passages such as these have inspired the Humean theory of motivation (from here onwards HTM) according to which an agent cannot be motivated by belief alone, but only by a belief-desire pair. More particularly, HTM states that an agent is motivated to act if and only if she ‘has an intrinsic desire for the world to be a certain way and a belief that her acting in the relevant way, a way which represents an option available to her, will result in the world’s being the way she intrinsically desires it to be’.21 Humeans about motivation thereby claim that their theory is presupposed by ‘all of the other explanations that we commonsensically give’,22 maintaining that the only difference between actions and (mere) bodily movements is that the former may always be explained in Humean terms that reveal the agent’s intention.23 Hume’s science of humanity, outlined in section 1 above, gives a far more central role to character than HTM might have us imagine.24 Indeed, Hume’s historical explanations are so unlike those produced by HTM that it would be charitable to question whether the latter is really to be found at all in Hume’s philosophical work. One way of resolving this question is that of examining specific explanations that Hume offers in The History of England to see whether they can easily be reconstructed into a Humean story. Baier does just this, reaching a sceptical conclusion.25 But even if the results had been positive, this would lend support to HTM

204

9780826443595_Ch09_Final_txt_print.indd 204

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS itself rather than the supposition that Hume held such a theory. One might try to argue that there is simply a deep inconsistency between Hume’s practical work as a historian and his philosophical theories, one which reflects his own meta-philosophical outlook (THN 1.4.7.2 / 264). But this would be premature for, as we have already seen, Hume reflects on the nature of human action in his writings on historiography. Moreover, none of his philosophical views lend any direct support to HTM; Hume scholars (as recently as Radcliffe, 1999)26 are wrong to suggest that he claimed that beliefs could never motivate alone. For one, the passages quoted at the start of this section are a red herring, the term ‘belief’ being conspicuous by its absence. More to the point, it is clear from his writings that Hume does not equate belief with reason. He uses the term ‘reason’ in a number of interrelated senses, describing it as a faculty of discovery (e.g. THN 3.1.1.9 / 458; cf. EHU 4.7 / 28), an instinct (e.g. THN 1.3.16.9 / 179), an equivalent to the general properties of the imagination (THN 1.4.7.7 / 267) and ‘an affectation of the very same kind as passion’ (THN 2.3.8.13 / 437).27 The last quotation derives from a passage in which he makes the following subtle distinction between reason and passion: What we commonly understand by passion is a violent and sensible emotion of mind, when any good or evil is presented, or any object, which, by the original formation of our faculties, is fitted to excite an appetite. By reason we mean affectations of the very same kind with the former; but such as operate more calmly, and cause no disorder in the temper: Which tranquillity leads us into a mistake concerning them, and

causes us to regard them as conclusions only of our intellectual faculties (THN 2.3.8.13 / 437). In other places reason is contrasted with experience (e.g. EHU 4.7 / 28), sentiment (e.g. EPM 1.3 / 170), and imagination (e.g. THN 1.3.9.19n22 / 117n1; cf. EHU 5.12 / 49), but never desire. Reason, it would seem, is too calm an affectation to be called an emotion, but hardly a product of the intellectual faculties. It might be better, then, to view reason as a specific (but by no means the only) source of belief, namely one capable of discoveries, as limited by Hume’s fork: Reason is the discovery of truth or falsehood. Truth or falsehood consists in an agreement or disagreement either to the real relations of ideas, or to real existence and matter of fact.28 Whatever, therefore, is not susceptible of this agreement or disagreement, is incapable of being true or false, and can never be an object of our reason . . . our passions, volitions, and actions . . . are original facts and realities, compleat in themselves, and implying no reference to other passions, volitions, and actions. ’Tis impossible, therefore, they can be pronounced either true or false, and be either contrary or conformable to reason (THN 3.1.1.9 / 458; cf. THN 2.3.3 / 413–8).29 Such discoveries may result in belief, but they are not its only source. Mutatis mutandis, not all beliefs are inert in Hume’s view. After all, he believes that ‘any thing may produce any thing’ (THN 1.3.15.1 / 173). At most, it is only those beliefs reached through reason alone that cannot motivate.30 But even this will prove to be a misleading way of putting things, since

205

9780826443595_Ch09_Final_txt_print.indd 205

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS (as we shall see) Hume cannot even allow that reason can alone produce beliefs of any kind. According to Hume, beliefs and/or opinions are lively ideas: ‘An opinion or belief is nothing but a strong and lively idea deriv’d from a present impression related to it.’ (THN 1.3.8.16 / 119; cf. THN 1.1.1.1 / 1–2; 1.3.7.5 / 96) Given that ideas and impressions differ only in their degree of vivacity, Hume naturally supposes that lively ideas such as beliefs exhibit the same effects as impressions, to an appropriately fainter degree (cf. THN 1.3.5.7 / 8). In fact, without the influence of belief all of our actions would be at the complete mercy of our impressions.31 In a section entitled ‘Of the Influence of Belief’ he writes: [T]he ideas of those objects, which we believe either are or will be existent, produce in a lesser degree the same effect with those impressions, which are immediately present to the senses and perception. The effect, then, of belief is to raise up a simple idea to an equality with our impressions, and bestow on it a like influence on the passions (THN 1.3.10.3 / 119). The context makes it clear that by ‘like influence’ Hume means ‘brings about the same effect to a lesser degree’, the degree in question being proportionate to the degree to which the idea in question is fainter to the impression of which it is a copy, beliefs being the most lively of all ideas. As Annette Baier has recently put it, ‘Hume does not exactly subscribe to a “belief + desire” analysis of motivation, since desires are only among the passions and sentiments which lead to action, and for him a main role for belief is to cause passions, as well as to instruct us on how to satisfy them’.32 In addition, while Hume contrasts reason with sentiment he explicitly identifies belief with it:

. . . belief is more properly an act of the sensitive, than of the cogitative part of our natures (THN 1.4.1.8 / 183). Belief is nothing but a peculiar feeling . . . or sentiment. . . . ’Tis felt rather than conceived, and approaches the impression, from which it is deriv’d, in its force and influence. (THN App. 3–9 / 624–7) So conceived, belief is an act of mind consisting of a ‘strong and steady conception of any idea’ (THN 1.3.7.5n20 / 96–7n1). To conceive of an idea in such a way, one which ‘approaches in some measure to an immediate impression’, is to be ‘perswaded of the truth of what we conceive’ (ibid.). Such persuasion does not merely accompany the simple conception (that would render the persuasion equivalent to an impression), but is a modification of it into something firmer. Hume felt that the account of belief outlined in the Treatise had been misunderstood, dedicating the first half of his Appendix to clarifying his notion of belief (whose previous expressions had ‘not been so well chosen’) in the hope of ‘guarding against all mistakes in readers’ (THN App.1 / 623). Here he distinguishes more explicitly between simple conceptions (viz. ideas) and firm conceptions (viz. the feeling that an idea is true). Beliefs are neither new ideas, nor impressions accompanying simple conceptions but, rather, firm conceptions of the very same idea which one may have previously conceived simply: if belief consisted merely in a new idea, annex’d to the conception, it wou’d be in a man’s power to believe what he pleas’d. We may, therefore, conclude, that belief consists merely in a certain feeling or sentiment . . . When we are convinc’d of any matter of fact, we do not but conceive

206

9780826443595_Ch09_Final_txt_print.indd 206

2/9/2012 9:35:54 PM

ACTION, REASON AND THE PASSIONS it, along with a certain feeling, different from what attends the mere reveries of the imagination. And when we express our incredulity concerning any fact, we mean, that the arguments for that fact produce not that feeling. Did not the belief consist in a sentiment different from our mere conception, whatever objects were presented by the wildest imagination, wou’d be on an equal footing with the most establish’d truths founded on history and experience. There is nothing but the feeling, or sentiment, to distinguish the one from the other . . . there is a greater firmness and solidarity in the conceptions, which are the objects of conviction and assurance, than in the loose and indolent reveries of a castle-builder . . . they strike upon us with more force; they are more present to us; the mind has a firmer hold of them, and is more actuated and mov’d by them . . . In short, they approach nearer to the impressions, which are immediately present to us. (THN App. 2–3 / 623–5) Rejecting the view that a simple belief might be ‘annex’d’ to a conception without modifying it ‘after the manner that will and desire are annex’d to particular conceptions of good and pleasure’ (THN App. 4 / 625), Hume concludes that what distinguishes beliefs from simple conceptions is the presence of some feeling or sentiment. Simply to conceive something is not to hold it true, but merely to have a possible truth present in one’s mind. By contrast, to believe that x is true is to feel that it is true. Tito Magri puts it nicely when he writes that to believe is ‘to have an idea present to the mind as if it were an impression’.33 By so modifying simple conceptions beliefs have the power to influence action: ‘The effects of belief, in influencing the passions and imagination, can all be explain’d from the firm conception.’ (THN

App. 7 / 626) What moves us to act, then, is not a simple conception but a belief. Reason alone cannot produce such a feeling. It may give rise to conceptions of matters of fact or the relation of ideas, but it cannot produce belief, let alone passion or action.34 A judgement may result in either knowledge or (mere) opinion or belief, but reason alone cannot cause it to do so, which is not to say that it cannot play an important role in the production of our beliefs and actions.35 Just as Lewis Carroll was later to demonstrate that entailment and inference are not the same thing,36 so Hume shows that inferential judgement does not amount to a belief; in each case the latter may result from the former, but it need not do so (nor ought it to, Hume would say). Pari Passu, one can judge that the truth of q follows from the truth of p, without coming to form the belief that q is true, even if one believes that p. Conversely, one can believe q to be true, without judging that its truth follows from that of anything else one believes. Such possibilities underlie the compatibility of sceptical judgements with non-sceptical beliefs (and vice versa, albeit less frequently). This is the sense in which human nature is stronger than reason.37 Hume goes to great lengths to show that this does not make our beliefs unreasonable.38 On the contrary, what is unreasonable is the thought that all our beliefs are the conclusions of reason alone. The above all seems to suggest that Hume does not equate beliefs and opinions with judgements. Unlike the latter, the former are not judged or conceived but felt. More to the point, while beliefs can alone cause action, judgements can only do so in combination with a passion:39 The action may cause a judgement, or may be obliquely caused by one, when

207

9780826443595_Ch09_Final_txt_print.indd 207

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS the judgement concurs with a passion . . . reason, in a strict and philosophical sense, can have an influence on our conduct only after two ways: Either when it excites a passion by informing us of the existence of something which is a proper object of it; or when it discovers the connexions of causes and effects, so as to afford us means of exerting any passion. These are the only kinds of judgement, which can accompany our actions, or can be said to produce them in any manner; and it must be allowed that these judgements may often be false and erroneous. (THN 3.1.1.11–12 / 459)

and Norton parse ‘ideas of the judgement’ as ‘ideas believed’,42 but this is somewhat rash. The only reason to suppose that Hume might use the words ‘judgement’ and ‘belief’ interchangeably is that he does not make a song and dance about their differences. In fact, he never offers a proper account of judgement,43 thereby forcing the reader either (a) to assume it is to be identified with belief or (b) to reconstruct the notion out of Hume’s distinction between simple and firm conceptions of ideas. Judging by his footnotes and appended clarifications, it is certainly plausible that judgements fall somewhere between the two:

Hume may not be a Humean about motivation but, pace Korsgaard, he does not quite maintain that ‘an action essentially is nothing more than a movement caused by a judgement or idea that regularly has an effect on the will’,40 for no idea could have an effect on the will unless it was sufficiently vivid to qualify as a belief and beliefs do not arise from pure reason but are typically explained by our character.41 On the other hand, it is worth recalling that beliefs, being ideas rather than impressions, are not ‘compleat in themselves’, thus remaining susceptible to the cogitations of reasons and capable of truth or falsehood; though, here as elsewhere, things are not helped by the fact that Hume makes no attempt to distinguish between one’s believing something and what one believes. To complicate things more, recall that for Hume, reason is itself an affection, differing from passion only in its degree of tranquillity. He accordingly also characterizes the ‘ideas of the judgement’ as he does sentiments: ‘it is something felt by the mind which distinguishes the ideas of the judgement from the fictions of the imagination . . . and renders them the governing principles of all our actions’ (THN 1.3.7.7 / 97). Norton

The error consists in the vulgar division of the acts of the understanding into conception, judgement, and reasoning, and in the definitions we give of them. Conception is defined to be the simple survey of one or more ideas; Judgement to be the separating or uniting of different ideas: Reasoning to be the separating or uniting of different ideas by the interposition of others, which show the relation they bear to each other. But these distinctions and definitions are faulty in very considerable articles . . . these three acts of the understanding . . . all resolve themselves into the first, and are nothing but particular ways of conceiving our objects. (THN 1.3.7.5n20 / 96–7n1) . . . the mind has a firmer hold, or more steady conception of what it takes to be a matter of fact, than of fictions (THN App. 5 / 626). A judgement is neither a simple conception nor a feeling or sentiment but the thought that something is true. Hume tells us precious little about what it is to judge that something is a matter of fact but the conception in question is arguably more vivid than imagining and

208

9780826443595_Ch09_Final_txt_print.indd 208

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS fainter than belief. Such an outlook would not only allow for the possibility (valuable to Hume’s scepticism) of judging that something is the case without believing it to be so, but it would also help to explain his controversial account of human morals. Numerous books and articles have been devoted to Hume’s account of the nature of moral judgement44 yet he never actually mentions moral judgement and, if anything like the picture outlined above is correct, he takes morals to be not judgements but beliefs. For morals, like beliefs but unlike judgements, have great motivational influence: ‘Morals excite passions and produce or prevent actions. Reason of itself is utterly impotent in this matter. The rules of morality, therefore, are not conclusions of our reason.’ (THN 3.1.1.6 / 457) This allows Hume to assert that it is more correct to speak of moral sentiments than of moral judgements: Morality, therefore, is more properly felt than judg’d of; though this feeling or sentiment is commonly so soft and gentle, that we are apt to confound it with an idea, according to our common custom of taking all things for the same, which have any near semblance to each other . . . To have a sense of virtue is nothing but to feel a satisfaction of a particular kind from the contemplation of character. The very feeling constitutes our praise and admiration . . . We do not infer a character to be virtuous because it pleases: But in feeling that it pleases after such a particular manner, we in effect feel that it is virtuous (THN 3.1.2.1–3 / 470–1). Given his idiosyncratic philosophy of mind, the question of whether or not Hume is a noncognitivist is fatefully anachronistic. What is clear is that there are no textual reasons

to suppose that when he states that belief includes a motivating capacity, he is working with a different notion of belief to that explored so far. Nor is he interested in metaethical questions about the nature of morality. Rather, Hume’s investigation focuses on how we come to reach the moral opinions and persuasions that we each have. His answer is that this is not a matter of discovery (be it through intuition or demonstration), but of feeling or sentiment. Commentators have equated moral sentiments with indirect passions,45 but this cannot be right, for the moral sentiments are ideas whereas all passions are impressions of a particular kind.

3. RULING PASSIONS AND THE WILL Hume presents his theory of the passions in Book II of the Treatise, eventually transformed into A Dissertation of the Passions (1757; originally published as an essay ‘of the passions’). The former divides into parts on (i) pride and humility, (ii) love and hatred (including benevolence, anger, malice, envy and lust), and (iii) the relation of passions to the will. Passions, for Hume, are secondary impressions of reflection as opposed to original impressions of sensation such as bodily pain and pleasure (THN 2.1.1.1 / 275). Secondary impressions arise either from original impressions of immediate sense-perception or from their ideas, for example, the memory of past sensation or the expectation of a future one. I may feel sad because I directly perceive something distressing or, just as often, because I recall (or merely believe or imagine) this had been the case. Both direct and indirect passions are ‘founded on pain and pleasure’ (THN 2.3.9.1 / 438). The former

209

9780826443595_Ch09_Final_txt_print.indd 209

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS require only this cause, whilst the latter also require a related object. Hume’s notion of a reflective impression is inspired by the Hellenistic thought that emotions may contain – or be closely related to – a cognition that is not discoverable by reason, though his own conservative stance is that ‘passions can be contrary to reason only so far as they are accompany’d with some judgement or opinion’, a thought which swiftly leads to the infamous remark about it not being contrary to reason ‘to prefer the destruction of the whole world to the scratching of my finger’ (THN 2.3.3.6 / 416). Hume further divides such passions into calm and violent ones. The former are said to include ‘the sense of beauty and deformity in action, composition, and external objects’. By contrast ‘the passions of love and hatred, grief and joy, pride and humility’ are all of the latter, violent, kind (THN 2.1.1.3 / 276). He is careful, however, to note that this ‘vulgar and specious’ division ‘is far from being exact’, noting that ‘the raptures of poetry and music frequently rise to the greatest height’ while ‘other impressions, properly call’d passions, may decay into so soft an emotion, as to become, in a manner, imperceptible’ (ibid.). Hume’s final division is between direct passions, which ‘arise immediately from good or evil . . . pain or pleasure’ (THN 2.1.2.3 / 276) and indirect passions which ‘proceed from the same principles, but with the conjunction of other qualities’ (ibid.). Parts I and II of Book II of the Treatise, and much of the Dissertation on the Passions, focus on the latter. These passions involve a reciprocal relation between sentiments and ideas and include such vices and virtues as pride, humility, ambition, vanity, love, hatred, envy, pity, malice and generosity (though Hume sees passions as the effects of vice and

virtue, see, for example, THN 2.1.7.2 / 295). The third part relates the will to direct passions, it being somewhat of a puzzle why Hume does not also allow for indirect passions to influence action on their own. The answer lies in his notion that a person desires to act (or omit from acting) in relation to perceived good and evil (THN 2.3.9.7 / 439) which he seems to equate with pleasure and pain (THN 2.3.3.3 / 414, 3.1.1.12 / 459).46 As Rachel Cohon persuasively argues, indirect passions cannot be motives to the will because they are not expectations of pleasure or pain but, rather, evaluative responses of people that do not directly relate to ‘the good or the absence of the evil that may be attain’d by any action of the mind or body’ (THN 2.3.9.7 / 439).47 If so, then not all moral sentiments are motives to the will either (it being the very same evaluative features that make them easy to conflate with certain indirect passions). Direct passions, such as desire, aversion, grief, joy, hope, fear, despair and security, are akin to – yet distinct from – the will. Whilst Hume labels the will the most remarkable of the ‘immediate effects of pain and pleasure’, it remains a mere effect and possesses none of the remarkable causal qualities that philosophers have traditionally associated with it: ‘by the will, I mean nothing but the internal impression we feel and are conscious of, when we knowingly give rise to a new motion of our body, or perception of our mind’ (THN 2.3.1.2 / 399). We have already seen that Hume believes that the will cannot be moved by reason alone, without the assistance of sentiments or passions. Of the influencing motives of the will, the greatest are the violent passions, although it remains true that ‘the calm ones, when corroborated by reflection, and seconded by resolution, are able to controul them in their most furious

210

9780826443595_Ch09_Final_txt_print.indd 210

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS movements’ (THN 2.3.8.13 / 437–8). As with all other aspects of the nature of will and the direct passions, these causes are the same in animals as they are in humans (see THN 2.3.9.32 / 448). Hume’s account of the virtues is Aristotelian insofar as it recognizes, pace Stoicism, that morality is largely a matter of having the right passions, at the right time, and to the right degree. His principle of association entails that certain impressions will invoke particular passions, be they direct or indirect, and that there may also be associations between passions of either kind.48 Whilst this does not guarantee a unity of the virtues, it suggests that they are very closely connected. In the Treatise, Hume declares that the principle of sympathy is ‘the chief source of moral distinctions’ (THN 3.3.6.1 / 618), allowing us to ‘enter into the sentiments of the rich and poor, and partake of their pleasure and uneasiness’ (THN 2.2.5.14 / 362). However, the principle plays no comparable role in the Enquiry concerning the Principles of Morals (1751), a drastically modified version of Book III of the Treatise, which places a greater emphasis on the sentiment of approbation.

3

4

5

6 7

8

9

NOTES 1

2

See A. Baier, Death and Character: Further Reflections on Hume (Cambridge, MA: Harvard University Press, 2008), p. 230 for the trouble this creates in relation to both of the Treatise definitions of a ‘cause’. Baier also notes that Don Garrett and Peter Millican have suggested that the dropping of the principle of contiguity from the definitions in the Enquiry might have occurred to allow for mental causation. Its absence would also seem to allow for action at a distance. See also P. F. Strawson, Scepticism and Naturalism (London: Routledge, 1985),

10

11

12

pp. 11ff.; D. Garrett, ‘Reasons to Act and Believe: Naturalism and Rational Justification in Hume’s Philosophical Project’, Philosophical Studies, 132(1) (2007), pp. 1–16, pp. 4–5; C. Sandis, ‘Hume and the Debate on “Motivating Reasons”’, in C. Pigden (ed.), Hume on Motivation and Virtue (Basingstoke: Palgrave Macmillan, 2009), pp. 142–54, pp. 151–2; and C. Sandis, ‘Induction’, in J. Hyman and H-J. Glock (eds), A Companion to Wittgenstein (Oxford: Wiley-Blackwell, forthcoming 2012). Hume does not use the word ‘event’ often, least of all in relation to actions. Also, it is not implausible to suggest that he often uses it in a sense that would make it interchangeable with ‘fact’ (cf. J.L. Austin, Philosophical Papers, Oxford: Clarendon Press, 1961, p. 156). See also C. Sandis, ‘A Just Medium: Empathy and Detachment in Historical Understanding’, in Journal of the Philosophy of History 5(1) (2011), pp. 179–200. Hume’s notion of sentiment being weak enough to be embraced by Stoicism, which he contrasts with apathy (E 151). See Baier, 2008, Death and Character, pp. 4–5. For a helpful exploration of Hume’s notion of character see P. Russell, Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility (Oxford: Oxford University Press, 1995), pp. 95ff. D. Knowles, The Historian and Character (Cambridge: Cambridge University Press, 1963), p. 10. One way out would be to follow Christine Korsgaard’s suggestion that perhaps Hume’s ‘notion of the person as the object of pride or love is not the same as the notion of the person as a bundle of successive perceptions’ (C. Korsgaard, The Constitution of Agency: Essays on Practical Reason and Moral Psychology (Oxford: Oxford University Press, 2008), p. 290). Cf. B. Croce, History as the Story of Liberty, trans. S. Sprigge (London: Allen & Unwin, 1941), p. 47, and Knowles, The Historian and Character, pp. 4–5; both quoted in E.H. Carr, What is History? (London: Penguin, 1961), p. 77. K.R. Merrill, Historical Dictionary of Hume’s Philosophy (Lanham, MD: The Scarecrow Press, 2008), p. 138. I am thinking here of deviant causal chains, whose challenge Davidson conceded to be

211

9780826443595_Ch09_Final_txt_print.indd 211

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS

13 14 15

16

17 18

19

20

21

22

23 24

problematic (see, for example, D. Davidson, Essays on Actions and Events (Oxford: Clarendon Press, 1980), p. 87). Baier, Death and Character, p. 12. Korsgaard, The Constitution of Agency, p. 292. See W. Sellars, ‘Volitions Reaffirmed’, in M. Brand and D. Walton (eds), Action Theory (Dordrecht: D. Reidel), pp. 47–63 and Davidson, Essays on Actions and Events, p. 65. For a helpful overview see H. Beebee, Hume on Causation (London: Routledge, 2006). My own view is defended in ‘Hume on the Meaning of “Necessity”’, in K. Allen and T. Stoneham (eds), Causation and Modern Philosophy (London: Routledge, 2011), pp. 166-87. Cf. Baier, Death and Character, pp. 226–7. See further below for whether Hume actually takes morals to be judgements. Hume never talks of motivation per se, but only ‘motives’ which he takes to ‘produce’ or ‘influence’ action. Contemporary technical jargon is Humean in so far as motivation is understood as a causal notion, but the motivation (or influence) of action is different from its production; we are frequently motivated to perform actions that never take place. (See C. Sandis, ‘Gods and Mental States: The Causation of Action in Ancient Tragedy and Modern Philosophy of Mind’, in New Essays on the Explanation of Action (Basingstoke: Palgrave Macmillan, 2008), pp. 356–83). Note that, unlike a number of his interpreters, Hume only ever capitalizes the word ‘reason’ when beginning a new sentence. M. Smith, ‘Humeanism about Motivation’, in T. O’Connor and C. Sandis (eds), A Companion to the Philosophy of Action (Oxford: Wiley-Blackwell, 2010), pp. 153–8, p. 153; cf. his The Moral Problem (Oxford: Blackwell, 1994), pp. 12, 92ff. M. Smith, ‘The Possibility of Philosophy of Action’, repr. in his Ethics and the A Priori (Cambridge: Cambridge University Press, 2004), p.157. Cf. Davidson, On Actions and Events, pp. 7–8. Cf. Baier, Death and Character; ‘Hume [on Action]’ in T. O’Connor and C. Sandis (eds), A Companion to the Philosophy of Action (Oxford: Wiley-Blackwell, 2010), pp. 513–20;

25

26

27

28

29

30

31 32

33

34

and Sandis, ‘Hume and the Debate on “Motivating Reasons”’, pp. 142–54. See Baier, Death and Character and ‘Hume [on Action]’. E. Radcliffe, ‘Hume on the Generation of Motives: Why Beliefs Alone Never Motivate’, Hume Studies 25(1) (1999), pp. 101–22. David Owen writes that ‘Locke is happy to use the same term for a faculty, the characteristic activity of that faculty, and the result of that faculty’ (Hume’s Reason (Oxford: Oxford University Press, 1999), p. 48). The same might be said of Hume’s use of terms like ‘reason’, ‘judgement’ and ‘passion’, although, as Owen points out, reason is, for Hume, at most a ‘sub-class of the imagination’, and even this characterization is problematically loose (ibid., pp. 75–6). Hume notes in his Appendix that ‘an inference concerning a matter of fact is nothing but the idea of an object, that has been frequently conjoin’d, or is associated with a present impression’ (THN App. 6 / 626). As shall become evident in section 3, this is not obviously true of the reflective impressions, particularly passions that are indirect (cf. N. Kemp Smith (The Philosophy of David Hume, London: Macmillan, 1941), p. 166). See C. Pigden, ‘If not Non-Cognitivism, then What?’ in his Hume on Motivation and Virtue, pp. 80–104 and Sandis, ‘Hume and the Debate on “Motivating Reasons”’. See also Owen, Hume’s Reason, p. 165. Baier, ‘Hume [on Action]’, pp. 514–15. Cf. C. Korsgaard, Self-Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009), p. 64n6. T. Magri, ‘Hume on the Direct Passions and Motivation’, in E.S. Radcliffe (ed.), A Companion to Hume (Oxford: Blackwell, 2008), pp. 185–200, p. 191. It is worth noting that this is not a view about the ontology of so-called motivating reasons. Contrary to what is assumed by both sides of the contemporary debate between psychologistic and non-psychologistic accounts of so-called ‘motivating reasons’, the consideration I act upon, and the belief that motivates me to act upon it, are not one and the same thing (for a detailed argument see Sandis, ‘Hume and the Debate on

212

9780826443595_Ch09_Final_txt_print.indd 212

2/9/2012 9:35:55 PM

ACTION, REASON AND THE PASSIONS

35

36

37

38 39

“Motivating Reasons”’; cf. M.M. Karlsson, ‘The Influencing Motives of the Will’, in S. Traiger (ed.), The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell, 2006), pp. 246–54). Karl Schafer (‘Review of C. Pigden (ed.), ‘Hume on Motivation and Virtue’, Notre Dame Philosophical Reviews, May 2010, sect. 1) has argued that Hume’s claim that reason is inert must be understood as a claim about the inability of the faculty itself to generate new volitions, passions or actions. The view presented here is sympathetic to this insight, but would add beliefs to the list of things that reason cannot produce alone. This is incompatible with Schafer’s claim that ‘beliefs about pleasure and pain’ may be the product of ‘abstract reasoning’ as well as his further suggestion that some other faculty (viz. a moral one) is required for motivation. C. Dodgson, ‘What the Tortoise Said to Achilles’, Mind 4 (1895), pp. 278–80. The early modern distinction between demonstrative and probabilistic reasoning (relating to knowledge and belief respectively) should not be conflated with the distinction between deductive and inductive reasoning (Owen, Hume’s Reasons, pp. 30ff., 83ff.). See Owen, Hume’s Reasons, pp. 144–6. This only amounts to the view that judgements cannot motivate alone if we conflate causation with motivation, and there is no reason to

40 41

42

43

44

45

46

47

48

suppose that Hume did so (see Sandis, ‘Gods and Mental States’ for why the conflation should be avoided). Korsgaard, Self-Constitution, pp. 63–4. We saw in section 1 that Korsgaard can be considerably more sensitive to this aspect of Hume’s account of action, though not persuaded by it. D.F. Norton and M.J. Norton, A Treatise of Human Nature – Oxford Philosophical Texts edn (Oxford: Oxford University Press, 2000), p. 454 (annotations to THN 1.3.7, note 7). See B. Stroud, ‘“Gilding or Staining” the World with “Sentiments” and “Phantasms”’, Hume Studies 19(2) (1993), pp. 253–72, p. 268. E.g. P. Foot, ‘Hume on Moral Judgement’, in D. Pears (ed.), Hume: A Symposium (London: Macmillan, 1963) and W. Brand, Hume’s Theory of Moral Judgement: A Study in the Unity of A Treatise of Human Nature (Dordrecht: Kluwer, 1992). See, for example, R. Cohon, ‘Hume’s Indirect Passions’, in E.S. Radcliffe (ed.), A Companion to Hume (Oxford: Wiley-Blackwell, 2008), pp. 160ff., 174–9. See also Karlson, ‘The Influencing Motives of the Will’, pp. 246–7. See Cohon, ‘Hume’s Indirect Passions’, pp. 172–3. See L. Alanen, ‘The Powers and Mechanisms of the Passions’, in Traiger, The Blackwell Guide to Hume’s Treatise, pp. 188–92 for detail.

213

9780826443595_Ch09_Final_txt_print.indd 213

2/9/2012 9:35:56 PM

10 FREE WILL James A. Harris

Hume made two contributions to what he and his time generally called the problem of liberty and necessity. The first was in Book Two of A Treatise of Human Nature (1739); the second was in Philosophical Essays concerning Human Understanding (1748), later retitled An Enquiry concerning Human Understanding. Much the same line of argument was presented in both places, but the change of context served to alter the larger significance of that line of argument. Both of Hume’s treatments of the problem of liberty and necessity will be considered here. The most obvious alteration that Hume made was one of tone: where, in the Treatise, he adopted an aggressive stance towards what at one point he termed ‘this fantastical system of liberty’ (THN 2.3.1.15 / 404), in the Enquiry he described his approach as a ‘reconciling project’ (EHU 8.23 / 95). It is thus the Enquiry’s treatment of the problem, in particular, that makes it plausible to understand Hume as an early exponent of what has come to be called ‘compatibilism’ with respect to moral responsibility. It should be kept in mind, however, that Hume’s primary concern is not with showing how the necessitation of human action might be reconciled with the legitimacy of holding human beings morally responsible for their actions. Much more prominent is an interest in action

as such, and whether or not there are reasons to regard human action as, in general, necessitated, and whether those reasons are also reasons to regard human action as unfree. These are taken by Hume to be questions concerning a matter of experiential fact. In this respect, if in few others, Hume was typical of his age, which had, for better or worse, little interest in distinguishing between ‘philosophical’ and ‘empirical’ questions. For Hume and most of his British contemporaries, the limits of human experience were the limits of intelligible thought and meaningful language. There was no realm in which a priori metaphysical investigations might be conducted. In his manner of dealing with the problem of liberty and necessity, Hume exploits the dependence of philosophy upon experience in an argument for the need of redefinitions of both freedom and necessitation. There was of course a considerable amount of discussion of the problem of liberty and necessity in the decades prior to the publication of the Treatise. Hobbes had debated the problem with Bishop Thomas Bramhall in the 1640s and 1650s.1 Locke had explored the nature of human freedom in the Essay concerning Human Understanding’s chapter ‘Of Power’, and had significantly altered his account in successive editions of the Essay,

214

9780826443595_Ch10_Final_txt_print.indd 214

2/9/2012 9:36:03 PM

FREE WILL though without making it clear exactly where he stood when it came to the question of whether liberty as he understood it is compatible with the necessitation of action. In 1717 the deist Anthony Collins had published A Philosophical Inquiry concerning Human Liberty, in which he adopted a markedly Hobbesian approach to the problem. Collins was in turn criticized by Samuel Clarke, he responded to the criticism, Clarke responded to the response, and a number of now obscure figures carried the debate on into the late 1720s.2 An unusual, strongly voluntarist, position had been taken up by the Irish Bishop William King in his De Origine Mali of 1702. King’s position had been critically examined by Leibniz in Essais de Théodicée (1717). Leibniz’s objections had been answered by Edmund Law in notes to his English translation of King, which was published in 1731. There is, however, no means of being sure that Hume read very much of this literature as he developed his own solution to the problem. He makes no explicit reference to any other author in the discussions of liberty and necessity in either the Treatise or the Enquiry. Nevertheless, he does seem to have approached the problem with previous solutions to it in mind. In the Abstract, the summary of the argument of Books One and Two of the Treatise that Hume published in order to boost sales in the spring of 1740, he claims that ‘what our author says concerning free-will’ ‘puts the whole controversy in a new light, by giving a new definition of necessity’ (Abs. 31 / 660, 34 / 661). It seems reasonable then to suppose that Hume intended to take the discussion of liberty and necessity in a new direction, and that he took himself to be doing something other than merely reiterating the kinds of ‘compatibilist’ arguments already proposed by Hobbes, Collins and Leibniz.

When one considers Book Two of the Treatise as a whole, it is not obvious what the function is of the two sections of Part III, ‘Of liberty and necessity’ and ‘The same subject continu’d’.3 Book Two is structured by a distinction between ‘indirect’ and ‘direct’ passions, and as Hume admits, the will is not, properly speaking, ‘comprehended among the passions’ (THN 2.3.1.1 / 399). Even so, he continues, a full understanding of the nature and properties of the will is necessary to the explanation of the passions. What he means by this is only apparent when one considers the larger argument of Book Two, and indeed also the larger argument of the first instalment of the Treatise taken as a whole.4 Having in Book One demolished the idea of reason as an autonomous faculty able to govern and control the economy of the human mind, Hume in Book Two presents an account of the realm of the passions as essentially self-regulating. The passions of each of us, that is to say, are regulated, by means of the operations of sympathy, by the opinions and feelings of those around us. Sociability, and with it the possibility of human society, is guaranteed by the extent to which, as Hume puts it earlier on in Book Two, ‘the minds of men are mirrors to one another’ (THN 2.2.5.21 / 365). We are acutely aware of how others, especially our equals and near-equals, see us, and the pains and pleasures arising out of that awareness prompt us to modify our passions, to make them less violent, and to acquire the selfcommand that on other accounts of the passions is supposed to be made possible only by means of the imposition of the order of reason upon the chaos of affection and emotion. Part III begins with what is in effect no more than a corollary of this view. The will is defined almost out of existence, as ‘nothing but the internal impression we feel and

215

9780826443595_Ch10_Final_txt_print.indd 215

2/9/2012 9:36:03 PM

FREE WILL are conscious of, when we knowingly give rise to any new motion of our body, or new perception of our mind’ (THN 2.3.1.2 / 399).5 The will, as traditionally understood, has no work to do in the theory of human nature that Hume has delineated in the earlier parts of the Treatise. By the same token, it has already been made obvious that human actions will be necessitated by the passions. Hume’s explicit argument for necessity forces the point home by showing that the doctrine of necessity, properly understood, neither has nor needs metaphysical or theological foundations. It has its warrant, rather, in everyday experience. Hume’s point of departure in both Treatise and Enquiry discussions of liberty and necessity is an examination of the operations of ‘body’, the material world which, so he says, is universally acknowledged to be necessitated in its every aspect. When, and only when, we know what we mean when we say that the actions of material objects are ‘necessary’ will we be in a position to know what it might mean to say that the actions of human beings are necessary as well. And, needless to say, Hume believes it is he who has at last made clear what it does mean to say that the actions of material objects are ‘necessary’. An understanding of his argument to the conclusion that human beings are subject to necessity depends intimately upon an understanding of his revisionist definition of causal necessity in general. The problem is that it is a matter of considerable debate what, exactly, Hume means to say about causal necessity in general. For present purposes, a host of contentious issues in this area must be ignored, and bald assertions will have to take the place of reasoned interpretative argument. Hume, it seems to the present writer, draws a distinction between causality as it might, or might not, operate deep in the

essence of material objects, and the causality that we have ideas of and speak about both in scientific pursuits and in ordinary life. Following Locke, he believes that we have no conception of the causal powers that things have in virtue of their essences. Essences – real essences, as distinct from nominal ones – lie outwith our cognitive grasp. The idea that we have of necessity is the product, not of perception of the operation of essential powers (for we have no such perception), but rather of associative propensities of the mind. Given sufficient exposure to constant conjunctions of types of events, those propensities issue in habits of prediction, and the feeling of confidence that goes along with those predictions is, according to Hume, the ‘impression’-source for the idea of causal necessity. What we really mean, therefore, when we call a connection between events ‘necessary’ and ‘causal’ (rather than accidental) is that there has been observed to be a regular conjunction of events of those types, and that we are disposed to make predictions as to future instances of the same conjunction in the future. That is all we mean. As Hume puts the point in the Treatise: ‘Here then are two particulars, which we are to consider as essential to necessity, viz. the constant union and the inference of the mind; and wherever we discover these we must acknowledge a necessity’ (THN 2.3.1.4 / 400). Hume proceeds to consider human actions in terms of these two ‘particulars’. First, he seeks to ‘prove from experience, that our actions have a constant union with our motives, tempers, and circumstances’ (THN 2.3.1.4 / 401). He says he believes that only ‘a very slight and general view of the common course of human affairs’ is needed for such a proof to be conclusive (THN 2.3.1.5 / 401). Hume’s confidence about the common course of human affairs is remarkable:

216

9780826443595_Ch10_Final_txt_print.indd 216

2/9/2012 9:36:03 PM

FREE WILL There is no light, in which we can take them, that does not confirm this principle. Whether we consider mankind according to the difference of sexes, ages, governments, conditions, or methods of education; the same uniformity and regular operation of natural principles are discernible. Like causes still produce like effects; in the same manner as in the mutual action of the elements and powers of nature. (THN 2.3.1.5 / 400) The point is not that all human beings in all times and places behave in the same way. It is rather that the various ways we have of distinguishing between people – in terms of sex, age, race, social rank – distinguish people into groups which are defined by constant and reliable patterns of behaviour. Hume provides no evidence for this claim. It is hard to believe that he means it to be taken as the result of anthropological and historical research that he has himself undertaken. He asserts the universal uniformity of human behaviour as if it were a truism. And to a considerable extent, that is indeed what it was at the time when Hume was writing – and for some time afterwards. Hume and his contemporaries were keenly interested in human diversity, across time as well as across space. They believed that human beings have a capacity for self-improvement, and that that capacity manifests itself in the many kinds of ‘progress’ in which eighteenthcentury writers, perhaps especially Scottish ones, interested themselves in displaying and explaining. But variety and progress could only be made intelligible, they all assumed, as the products of interactions between environment and a uniform human nature.6 Confidence in the existence of a uniform human nature is implicit in the very project (and title) of A Treatise of Human Nature, and at no point in the book does Hume give

an argument to provide that confidence with reasoned evidential support. It is in fact hard to imagine what such an argument might look like. Having shown that, as he puts it, ‘the union betwixt motives and actions has the same constancy, as that in any natural operation’, Hume moves on to the second ‘particular’ essential to necessity, proof that the influence of that union ‘is also the same, in determining us to infer the existence of the one from that of another’ (THN 2.3.1.14 / 404). Again, Hume’s view is that there is no doubt that this is so. His evidence is the general acceptance of ‘moral evidence’ in deliberations and decisions concerning human affairs. This was a semi-technical term of art, widely used in the seventeenth and eighteenth centuries, to denote a kind of probability arising from observed tendencies of human nature. When something was ‘morally’ certain, it was regarded as being extremely likely, so likely as not to admit of reasonable doubt, while at the same time not being so absolutely certain that doubt was impossible. Such certainty was taken to be all that we can pretend to in what we might learn from human words and actions – and was often contrasted with the perfect certainty we can have with respect to what is communicated to us directly by God. In this part of his argument Hume takes himself to be doing no more than reminding the reader of a familiar feature of human experience: that we generally are certain, albeit only ‘morally’ so, when it comes to what other people will do in the future, and that we manifest such certainty in all of the common affairs in life, in politics, in war, in business, in household affairs. For the most part we are not perturbed by the fact that it is never a matter of absolute certainty how another person will behave. We take advantage of our, and

217

9780826443595_Ch10_Final_txt_print.indd 217

2/9/2012 9:36:03 PM

FREE WILL others’, experience of the world regardless. And, Hume says, ‘whoever reasons after this manner, does ipso facto believe the actions of the will to arise from necessity, and that he knows not what he means, when he denies it’ (THN 2.3.1.15 / 405). It would appear that in this argument for the necessitation of human actions by motives, tempers and circumstances, Hume is being careful to stay, as it were, on the empirical surface of things, and to withhold all judgement about the metaphysical question of whether, in any one case, it is possible for someone to act otherwise than he or she in fact does. Here again there are complicated and vexed interpretative issues, but there is reason to believe that, in fact, Hume’s own doctrines, doctrines elaborated in Part III of Book One of the Treatise, make it impossible for Hume to be claiming anything stronger than that, as a matter of empirical fact, human behaviour is regular, and that, as another matter of empirical fact, we use that regularity as the basis for the assumptions and predictions we make in our dealings with other people. For Hume has argued in Part III of Book One that there is no reason to believe that it is impossible that the laws of nature, including of course laws of human nature, could not alter at some point in the future. Even if, as is not the case, we had knowledge of the powers by means of which causes necessitate their effects, we could not be certain that those powers might not differ in their operations from one moment to another. There is, therefore, no metaphysical straitjacket, so to speak, to ensure that the laws responsible for how things have happened in the past must obtain in all times and all places. The doctrine that, since the nineteenth century, has gone by the name of ‘determinism’ has, and can have, nothing other than experience to back it up, but the

truth is that experience does not, and cannot, back it up. Some philosophers of the early modern period, including Hobbes and Collins, argued that it is literally impossible for anything to have happened differently from how it did happen. They argued for this conclusion from the premise that, in the case of any and every event, it is impossible for an event not to have a cause, and from the further premise that a cause, in so far as it is a sufficient condition of the effect’s taking place, makes it impossible for the effect not to take place. Hume has rejected both of these premises. He has argued that there is no good reason to believe that every event must have a cause, and that it is always conceivable that a particular cause might not have its usual effect. If determinism involves confidence that there is something in the metaphysical order of things that ensures that the laws of nature cannot change, Hume is not a determinist. It was the fact that he had drained the doctrine of necessity of metaphysics that, it may be presumed, prompted him to claim, in the Abstract, that he had put the whole free will question in a new, purely empirical, light. As already mentioned, in the Treatise Hume writes as if his argument for the doctrine of necessity is at the same time an argument against the doctrine of liberty. ‘There is no philosopher,’ he writes, ‘whose judgement is so riveted to this fantastical system of liberty, as not to acknowledge the force of moral evidence, and both in speculation and practice proceed upon it, as upon a reasonable foundation’ (THN 2.3.1.15 / 399). He asserts also that to hold to the liberty of human action is to accept that human actions are uncaused, that therefore liberty ‘is the very same thing with chance’, and that, ‘[a]s chance is commonly thought to imply a contradiction, and is at least directly

218

9780826443595_Ch10_Final_txt_print.indd 218

2/9/2012 9:36:04 PM

FREE WILL contrary to experience, there are always the same arguments against liberty or free-will’ (THN 2.3.1.18 / 407). The claim that the idea that the will is free entails that the will is uncaused in its determinations was a common one among necessitarians in the early modern period. Libertarians replied that, on the contrary, choices and decisions were indeed caused, but by the agent him or herself, not by motive, temper or circumstance. Hume does not engage with that way of construing the doctrine of liberty. He imagines that the defender of liberty will have to rest his case on a conception of human actions as unpredictable because arbitrary. For reasons adumbrated in the previous paragraph, it was not in fact possible for Hume to claim that chance – in the form of an uncaused event – implies a contradiction. The weight of his case against ‘liberty or free-will’ lies in the claim that chance, or arbitrariness and randomness, is contrary to experience. He has already disarmed the objections of those who dwell upon the capriciousness and inconstancy of human beings. It is not always the case that we can explain and predict the behaviour of material objects, he has pointed out, but we do not conclude from that that those objects are not subject to causal necessity. There is a hint, however, in the second part of the Treatise discussion of liberty and necessity, that Hume is not so much arguing against liberty as seeking to redefine it. One of the reasons why people object to the doctrine of necessity, he observes, is the fact that they imagine that it implies ‘something of force, and violence, and constraint, of which we are not sensible’ (THN 2.3.2.1 / 407). Hume’s answer is that it is a mistake to imagine that necessity implies force, violence and constraint. The philosophy of ‘the schools’ – that is, the philosophy taught in universities through the late Middle Ages and on into

the early modern period, even into the early eighteenth century – makes a useful distinction between ‘the liberty of spontaneity’ and ‘the liberty of indifference’. The latter kind of liberty involves a negation of necessity and an equation of freedom with an absence of causation. But the former kind of liberty involves merely an absence of constraint and coercion, and is, therefore, as Hobbes, Locke and others had argued, compatible with the determination of action by motives. Hume does not note that others before him had argued that liberty in this sense is compatible with necessity. He says only that when people talk of liberty, they usually mean only an absence of constraint and coercion, and that this is the kind of freedom ‘which it concerns us to preserve’ (THN 2.3.2.1 / 407–8). A second reason why people tend to be repelled by the doctrine of necessity is to be found in what Hume terms ‘a false sensation or experience even of the liberty of indifference; which is regarded as an argument for its real existence’ (THN 2.3.2.2 / 408). By ‘false’ here Hume means deceptive. It was indeed a common claim of opponents of the doctrine of liberty that we have an introspective awareness of ourselves, both at the time of action and in retrospective reflection upon actions of the past, as radically free in our choices and decisions. When we are deliberating what to do, libertarians pointed out, it feels as if what will happen next is completely up to us. When we act on our deliberations, we have a sense of being undetermined and uncaused. When we look back on what we have done, we often have a feeling, and sometimes a very troubling one, of its having been possible for us to have acted differently from how we did. The eighteenth century, as was noted at the beginning of this chapter, was a self-consciously empirical, or ‘experimental’, period in philosophy, and it

219

9780826443595_Ch10_Final_txt_print.indd 219

2/9/2012 9:36:04 PM

FREE WILL was open to libertarians to ask why this vivid and common experience of freedom should be ignored in debates concerning liberty and necessity.7 Hume, in general, was not a philosopher especially attuned to or interested in experience from the first-person standpoint. In his moral philosophy, for example, his focus is resolutely upon the judgements we make of other people, and he has very little to say about judgements we make of ourselves. His theory of the passions, as we have seen, has at its heart the idea that we gain our sense of ourselves from awareness of how others see us. And when faced by what appears to be experience of a radical freedom to choose to act on any one of our motives, or indeed on no motive at all, Hume has no hesitation in judging that that experience is deceptive, and that the truth lies in the observer’s third-person perspective on human action. From that perspective, human behaviour looks regular and predictable. And such regularity and predictability license the conclusion that human behaviour is necessitated. The supposed first-person experience of a liberty of indifference can, therefore, be explained away as an illusion, of no significance when it comes to an examination of the experience relevant to making a decision about the doctrine of necessity. Later, more aggressive opponents of liberty, such as David Hartley and Joseph Priestley, were to go further and simply deny that there is any first-person experience of the freedom of the will. The third and final reason for the prevalence of the doctrine of liberty, Hume says, ‘proceeds from religion’ (THN 2.3.2.3 / 409). Religion teaches us that we are morally responsible for what we do, in the sense that our actions will be judged by God and used as a basis for either salvation or damnation. We are all under God’s law, and in order

for God’s law to be just, we must be free in how we act. We are of course also under human laws, and, for the same reasons, in order for those laws to be just, we must be free in how we act. Hume’s response is that, on the contrary, in order for laws, whether human or divine to be just, and also efficacious, our actions must be presumed to be determined by our motives and character. Consider the efficaciousness of law first. If the function of law is to prevent crime, then it must be assumed that laws of certain kinds will motivate people not to act in criminal ways, and that assumption contains the further assumption that human behaviour is regular and predictable. If it were not regular and predictable, law-makers could have no idea which laws to make, and no confidence that any laws could be relied upon to modify human behaviour. When it comes to the justness of the rewards and (more pertinently) the punishments that are decreed to go along with the keeping and violation of laws, it is necessary that there be some way of deciding whether a particular action was an accident or in some other way not one for which the agent is liable, or whether it was intended, caused by that intention, and in every other respect one for which the agent is genuinely liable. Hume’s position is that it is, in effect, the doctrine of necessity that is used to differentiate between liability and non-liability. We do not punish actions. We punish people in respect of their actions and, roughly speaking, we take people to be responsible for their actions to the extent that those actions are, as we say, in character. What it is for an action to be in character is, simply, for there to have been experience of that person, or that kind of person, doing that kind of thing before. It is, in other words, for the action to be subject to necessity as Hume has defined necessity. ‘Actions are by their very nature

220

9780826443595_Ch10_Final_txt_print.indd 220

2/9/2012 9:36:04 PM

FREE WILL temporary and perishing,’ Hume observes; ‘and where they proceed not from some cause in the characters and disposition of the person, who performed them, they infix not themselves upon him, and can neither redound to his honour, if good, nor infamy, if evil’ (THN 2.3.2.6 / 411). If the libertarian – where ‘libertarian’ means a defender of the liberty of indifference – were right, it would not be possible to attach actions to character, since every action is radically free, and proceeds from nothing durable or constant in the agent. ‘’Tis only upon the principles of necessity,’ Hume concludes, ‘that a person acquires any merit or demerit from his actions, however the common opinion may incline to the contrary’ (THN 2.3.2.6 / 411). In Book Three of the Treatise Hume returns to this line of thought and uses it to motivate one of the most controversial aspects of his moral philosophy: the claim that there is no difference in kind between ‘moral virtues’ and ‘natural abilities’ (such as intelligence and knowledge, wit and good humour, industry and perseverance).8 Those who insist upon a deep difference here tend to lay great stress upon the fact that natural abilities are, as Hume puts it, ‘entirely involuntary, and have therefore no merit attending them, as having no dependance on liberty and free-will’ (THN 3.3.4.3 / 608). Hume has three things to say in reply. First he points out that the picture of moral virtue common in ancient philosophy includes many things, including constancy, fortitude and magnanimity, which are not under the control of the will. Either, then, the ancients were wrong about the nature of virtue, or voluntariness is not the all-important feature of virtue that some people now say it is. Hume made this point knowing that many of his contemporaries took ancient thought just as seriously as he did, if not more so:

he was forcing them to choose between their respect for the ancients and their attachment to the will-based morality of Christianity. Secondly, Hume points out that if you accept that moral sentiments ‘arise from the natural distinctions of pain and pleasure’, you will struggle to make a principled difference between moral virtues and natural abilities, since it is plain that natural abilities excite pain and pleasure in the observer just as moral virtues do (THN 3.3.4.3 / 608–9). Here Hume has in his sights Hutcheson’s commitment to the distinction between virtues and abilities – though it was, of course, perfectly possible for the opponent of moral sense theory to see Hume’s argument as a reductio ad absurdum of the whole position. And: ‘Thirdly, as to free will, we have shown that it has no place with regard to the actions, no more than the qualities of men. It is not a just consequence, that what is voluntary is free. Our actions are more voluntary than our judgements; but we have not more liberty in the one than in the other’ (THN 3.3.4.3 / 609). Hume goes on to explain, in Mandevillean manner, how the distinction between the voluntary and involuntary has been used by ‘moralists’, and also by legislators and divines, as a basis for inventing this spurious distinction between the moral and the natural. The truth is that the latter distinction is a matter of words alone.9 Hume’s discussion of liberty and necessity in the Treatise is thus part of a larger argument concerning the place of the passions in human life, and concerning the basis of moral distinctions. His decision during the recasting of the doctrines of the Treatise into more digestible essay form to include ‘Of Liberty and Necessity’ in the new treatment of the understanding alters how one reads Hume on free will. He could have included it in either An Enquiry concerning the Principles

221

9780826443595_Ch10_Final_txt_print.indd 221

2/9/2012 9:36:04 PM

FREE WILL of Morals (1751) or the ‘Dissertation on the Passions’ (1757), but he did not, and so the question arises what place it has in the first Enquiry considered as a whole. In the first six sections of the Enquiry Hume strips his theory of the understanding down to its essentials, and presents a single line of argument concerning the nature of probabilistic reasoning. In the second six sections Hume concentrates on the sceptical consequences that line of argument has for a variety of controversial philosophical questions – in particular, questions concerning the nature of the necessary connection between causes and effects, the freedom of the will, the rational basis of revealed religion and the origins of the universe. He also seeks to clarify the precise nature of the scepticism he is advocating, and to make it clear that it is a purely speculative kind of scepticism, with no implications for everyday life. He describes the problem of liberty and necessity as ‘the most contentious question of metaphysics, the most contentious science’ (EHU 8.23 / 95); but the main aim of his discussion of the problem here in the Enquiry is to show that all that has been thought to be difficult when it comes to free will is the product of philosophical confusions and fallacies. His target is metaphysics and metaphysicians, and not, as in the Treatise, an aversion to the doctrine of necessity that is shared by ordinary people and philosophers alike. He is much more explicit in presenting the doctrine of necessity as something that we all already believe, did we but know it. He foregrounds the claim, present in the Treatise but not prominent, that our ordinary conception of liberty is perfectly compatible with the doctrine of necessity. In calling his approach to liberty and necessity a ‘reconciling project’, he is suggesting that once the claims of metaphysicians are shown to be groundless, the

whole dispute will cease, because there will be nothing left to argue about. The metaphysical claims that have bedevilled the free will problem are claims that are generally thought (by both advocates of necessity and their opponents) to be contained within the doctrine of necessity. These are claims concerning the existence and nature of powers that causes have, considered in themselves – powers that make effects happen, and that are common to both material and mental causes. The effect of juxtaposing ‘Of Liberty and Necessity’ with ‘Of the Idea of Necessary Connexion’ is to emphasize the fact that no such claims are made by the necessitarian who understands what our idea of necessity really amounts to. A prominent theme in ‘Of the Idea of Necessary Connexion’ is, as Hume puts it, ‘the weakness and narrow limits of human reason and capacity’ (EHU 7.28 / 76). In the context of the free will debate, the limits of human reason mean that the necessitarian claim that human actions are no different from the operations of material objects in their causal determination is robbed of all threatening or subversive implications. All that is being said in either case is that there is experience of constant conjunctions of types of events, and that that experience is taken as a ground for the making of inferences as to the causes and effects of observed events. A passage from the Treatise that is retained almost without alteration in the Enquiry acquires a particular importance in the later text: We may here be mistaken in asserting that there is no idea of any other necessity or connexion in the actions of body: But surely we ascribe nothing to the actions of the mind, but what everyone does, and must readily allow of. We change no circumstance in the received orthodox system with regard to the will, but only in that with regard to material

222

9780826443595_Ch10_Final_txt_print.indd 222

2/9/2012 9:36:04 PM

FREE WILL objects and causes. Nothing, therefore, can be more innocent, at least, than this doctrine. (EHU 8.27 / 97) It is the changes that Hume has made to the received orthodox system with regard to material objects and causes that enable his project of reconciliation. Hume dwells on the implications of these changes for the understanding of necessitarianism for rather more time than he spends on the doctrine of liberty. There is in this way a marked difference between Hume’s approach to the question of liberty and necessity, as restated in the first Enquiry, and, for example, Hobbes’s approach: where Hobbes had argued that what needed to be revised was our understanding of liberty, Hume holds that it is our – that is, philosophers’ – understanding of necessity that has to alter. The tone of the treatment of the free will question in the Enquiry is set when Hume declares early on that he hopes ‘to make it appear that all men have ever agreed in the doctrine both of necessity and of liberty, according to any reasonable sense, which can be put on these terms; and that the whole controversy has hitherto turned merely upon words’ (EHU 8.4 / 81). With respect to necessity, the argument proceeds in the same manner as the argument of the Treatise: first necessity as usually applied to the operations of matter is defined, in terms of regularities and inferences made on the basis of those regularities; then it is shown that the same kinds of regularities and inferences are to be found in the realm of human behaviour. There does appear to be a slight change of tack, however: what Hume is concerned to show is that it is generally believed that human behaviour is regular and predictable, and there is less of a sense that Hume himself holds to the doctrine of the uniformity

of human nature as a matter of unexamined dogma. He spends more time, too, dealing with the worry that there are ‘some actions, which seem to have no regular connexion with any known motives, and are exceptions to all the measures of conduct which have been established for the government of men’ (EHU 8.12 / 86). In response, he develops a distinction between how ‘the vulgar’ think about the uncertainty of events and how ‘philosophers’ do, a distinction that is consonant with the rather more careful attention to the practice of natural philosophy that is characteristic of the first Enquiry in general.10 There is more attention also to the practices of historians and of what we might now call anthropologists and sociologists. It is perhaps a consequence of this that even less interest is shown in the deliverances of first-person introspection than in the Treatise: only in a footnote is mention made of ‘a false sensation or seeming experience which we have, or may have, of liberty or indifference, in many of our actions’ (EHU 8.22 / 94). And when Hume finally turns to showing that ‘all men have ever agreed in the doctrine of liberty as well as in that of necessity’ (EHU 8.23 / 95), he is clearer than he was in the Treatise, arguing straightforwardly that experience suggests that all we can mean by liberty is ‘a power of acting or not acting, according to the determinations of the will; that is, if we choose to remain at rest, we may; if we choose to move, we also may’; such liberty, he continues, taking his cue from Locke, ‘is universally allowed to belong to every one who is not a prisoner and in chains’ (EHU 8.23 / 95). A further significant difference between the Treatise and Enquiry accounts is in what is said concerning the interest of religion in the question of liberty and necessity. In the Enquiry, but not in the Treatise, Hume acknowledges the fact that the doctrine of

223

9780826443595_Ch10_Final_txt_print.indd 223

2/9/2012 9:36:04 PM

FREE WILL necessity had appeared to many to have the consequence that it is God, and not individual human agents, who is responsible for the moral evil that there is in the world. The only alternative to this conclusion, as Hume mentions, had been taken to be a denial of the existence of moral evil, and the concomitant claim that evil is a matter of appearance only, and that the truth of the matter is that, as Alexander Pope had put it in the Essay on Man, ‘whatever is, is right’. But, it had been said by some, both conclusions are obviously absurd: God cannot be both good and a doer of evil; and the reality of moral evil is too palpable to be denied. And therefore, the same people had concluded, the doctrine of necessity is shown to be false. Hume makes it very plain that he agrees with those who find it impossible to deny the existence of moral evil. Just as physical ills, in the form of pain and suffering, impress themselves on the mind with a potency that makes a mockery of the claims of those who believe, or say they believe, that there is goodness to everything in the order of nature, so also moral harm generates painful feelings of outrage and resentment that are not diminished by reflection upon the possible contribution of crimes to the good of the universe considered as a whole. ‘Are such remote and uncertain speculations,’ Hume asks, ‘able to counterbalance the sentiments which arise from the natural and immediate view of the objects?’ (EHU 8.35 / 102). The reality of the difference between good and evil is ‘founded in the natural sentiments of the human mind’, sentiments which ‘are not to be controuled or altered by any philosophical theory or speculation whatsoever’ (EHU 8.35 / 103). Hume introduces the problem of evil as a further element of the Enquiry’s attempt to limit the pretentions of speculative reason. Reason’s weakness makes impossible not only traditional metaphysics, but also

traditional rational religion. Hume ends the discussion by noting that it has been found to exceed the power of philosophy to reconcile the freedom of human action with theological notions of prescience and predestination. Philosophy, therefore, should ‘return, with suitable modesty, to her true and proper province, the examination of common life’ (EHU 8.36 / 103).11 Despite Hume’s having, in the Abstract, drawn attention to the novelty of his way of treating the question of liberty and necessity, it went for the most part unnoticed in what reviews the Treatise and the first Enquiry did receive. A review of Books One and Two of the Treatise in the periodical Common Sense in July 1740 acknowledged the ‘moral evidence’ of inferences concerning human behaviour, but asserted that, even so, ‘every Man must be convinced from what he feels within himself, that this Influence is not absolute and necessary; and Self-Conviction is a much stronger Proof than any we can have from our Observation of external Objects, because we cannot know their Tempers and Circumstances, and much less the Motives they are governed by, so well as we do our own’.12 Joseph Highmore, in his Essays, Moral, Religious and Miscellaneous (1766), claimed that in the Enquiry Hume ‘maintains absolute uncontroulable necessity, in the moral as well as the natural world’, and, like the reviewer in Common Sense, was particularly troubled by the tension between such necessity and the existence, which he took Hume to allow, of ‘an universal feeling of liberty, and free agency in man’.13 ‘Surely’, Highmore wrote, either man is free, feels and knows himself to be so, and thence arises remorse or self-blame, on his having acted against his judgement and conscience; - Or, he is absolutely necessary, finds and knows

224

9780826443595_Ch10_Final_txt_print.indd 224

2/9/2012 9:36:05 PM

FREE WILL himself to be so; and therefore cannot consistently blame himself, or suffer remorse, for any past action; - Or, lastly, it is a matter of doubt with him, whether he is free or not; in which state of mind, he cannot be wholly without anxiety, nor perhaps without some kind or degree of remorse.14

In 1795 John Allen, signing himself ‘a Necessitarian’, published a pamphlet defending Hume’s account of necessity in the Enquiry from criticisms made by James Gregory in his Philosophical and Literary Essays (1792). But it could not be said that Hume’s version of the doctrine of necessity was central to the free will debate in the second half of the eighteenth century. In his Essays on the Active Powers of Man (1788), for example, the libertarian Thomas Reid takes as his target Joseph Priestley’s The Doctrine of Philosophical Necessity Illustrated (1777). The dispute between Reid and Priestley determined the contours of disputations about human freedom far into the nineteenth century. The distinctive features of Hume’s approach to the question of liberty and necessity went ignored even as much of the rest of his philosophy received renewed attention in the second half of the twentieth century. This was because it was assumed that Hume said nothing on the topic that Hobbes and Locke had not said before him. That assumption met with sustained criticism in Paul Russell’s Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility, where it is argued that Hume’s reconciliation of freedom, in the form of moral responsibility, with necessity depends to a significant extent upon his naturalistic analysis of the moral sentiments.15 It will be obvious to the reader that the present author also disagrees with the view that Hume is merely restating

old arguments in his writings on liberty and necessity.16 It has been argued here that Hume can only be understood to be a ‘determinist’ in an extremely qualified sense.17 Recently John P. Wright and Peter Millican have criticized this view, and have reasserted the contrary view that Hume’s determinism is full-blooded and unequivocal.18 Another recent development in the study of Hume on free will has been that it has come to seem to some – understandably enough, given the way Hume inserted ‘Of Liberty and Necessity’ into the first Enquiry immediately after ‘Of the Idea of Necessary Connexion’ – that Hume’s way of treating the free will question might have some relevance to the so-called ‘New Hume debate’: that is, the debate as to whether Hume is a ‘regularity theorist’ about causation or, instead, a kind of realist about the existence of unobserved and unobservable causal powers.19 On the interpretation offered here, what Hume brings to both the question of causation and the question of liberty and necessity is a kind of agnosticism about fundamental metaphysical issues. All we have to go on when it comes to such questions, according to Hume, is experience; and what experience tells us is, as Hume puts it, ‘that we can draw inferences concerning human actions, and that those inferences are founded on the experienced union of like actions, with like motives, inclinations, and circumstances’ (EHU 8.27 / 97).20

NOTES 1

2

For a modern edition of representative samples of the Hobbes-Bramhall debate, see V. Chappell, Hobbes and Bramhall on Liberty and Necessity (Cambridge: Cambridge University Press, 1999). For an account of the Collins-Clarke debate, see J.A. Harris, Of Liberty and Necessity: The

225

9780826443595_Ch10_Final_txt_print.indd 225

2/9/2012 9:36:05 PM

FREE WILL

3

4

5

6

7

8

9

10

Free Will Debate in Eighteenth-Century British Philosophy (Oxford: Clarendon Press, 2005), chap. 2. For some suggestions, see P. Russell, The Riddle of Hume’s Treatise: Skepticism, Naturalism, and Irreligion (New York: Oxford University Press, 2008), chap. 16. For a fuller development of this reading of Books One and Two of the Treatise, see J.A. Harris, ‘“A compleat chain of reasoning”: Hume’s Project in A Treatise of Human Nature, Books One and Two’, Proceedings of the Aristotelian Society 109 (2009), pp. 129–48. For a useful account of Hume’s definition of the will, see R.F. Stalley, ‘The Will in Hume’s Treatise’, Journal of the History of Philosophy 24 (1986), pp. 41–53. See C.J. Berry, Social Theory of the Scottish Enlightenment (Edinburgh: Edinburgh University Press, 1997), chap. 3. The importance to the free will debate of firstperson experience is a major theme of Harris, Of Liberty and Necessity, passim. For a detailed consideration of this aspect of Hume’s argument, see P. Russell, Freedom and Moral Sentiment: Hume’s Way of Naturalizing Responsibility (New York: Oxford University Press), chaps 6–12. Users of the new Clarendon edition of the Treatise should be aware that, without any textual authority, David and Mary Norton change the title of Section Five of Part III of Book Three from ‘Some farther reflections concerning the natural virtues’ to ‘Some farther reflections concerning the natural abilities’. Hume intended ‘natural virtues’, it may be presumed, as part of his strategy of unsettling the reader’s grip on the distinction between virtues and abilities. For more on relevant differences between the first Enquiry and the Treatise, see P. Millican, ‘Hume’s Sceptical Doubts concerning Induction’, in P. Millican (ed.), Reading Hume on Human Understanding (Oxford: Oxford University Press, 2002), pp. 107–73, esp. pp. 144–5.

11

12

13 14 15

16

17

18

19

20

For further discussion of the religious dimension of Hume’s discussion of liberty and necessity see T. Holden, Spectres of False Divinity: Hume’s Moral Atheism (Oxford: Oxford University Press, 2010), chap. 7. This and many other contemporary responses to Hume on liberty and necessity are reprinted in J. Fieser (ed.), Early Responses to Hume’s Metaphysical and Epistemological Writings, 2 vols, 2nd rev. edn (Bristol: Thoemmes Continuum, 2005); for the passage quoted, see vol. 1, p. 89. Fieser (ed.), Early Responses, vol. 1, p. 172. Ibid., p. 173. See especially Russell, Freedom and Moral Sentiment, Introduction. For further development of the reading offered here, see Harris, Of Liberty and Necessity, chap. 3, and J.A. Harris, ‘Hume’s Reconciling Project and “the common distinction betwixt moral and physical necessity”’, British Journal for the History of Philosophy 11 (2003), pp. 451–71. For a similar reading, see D. Garrett, Cognition and Commitment in Hume’s Philosophy (New York: Oxford University Press, 1997), chap. 6. See J.P. Wright, Hume’s A Treatise of Human Nature: An Introduction (Cambridge: Cambridge University Press, 2009), chap. 5; and P. Millican, ‘Hume’s Determinism’, Canadian Journal of Philosophy 40 (2010), pp. 611–42. For this debate, see R. Read and K. Richman (eds), The New Hume Debate, rev. edn (London: Routledge, 2007). The relevance of Hume’s discussions of ‘liberty and necessity’ to the debate is stressed especially by P. Millican, ‘Hume, Causal Realism, and Causal Science’, Mind 118 (2009), pp. 647–712, sect. 8. I am grateful to Sean Greenberg for comments on a draft of this chapter.

226

9780826443595_Ch10_Final_txt_print.indd 226

2/9/2012 9:36:05 PM

11 HUME ON MIRACLES Duncan Pritchard and Alasdair Richmond

1. HUME’S ARGUMENT: READINGS AND MISREADINGS Hume’s seminal ‘Of Miracles’ (EHU 10 / 109–31) still attracts heated discussion and different interpretations of his argument persist. What is clear, however, is that he held that rational beliefs in miracles based purely on testimony are (at very least) highly problematic. In particular, anyone whose belief in a particular religious hypothesis is due purely to testimony recorded hitherto that a miracle has occurred (‘miracle-testimony’) has failed to form their beliefs as Hume prescribes. Despite its enduring popularity as an anthology piece, ‘Of Miracles’ is not a standalone work but one closely tied to Hume’s overall philosophical project. In particular, ‘Of Miracles’ should be considered against the background of his regularity theory of induction. He famously did not believe that expectations that the future will resemble the past (or that the unobserved must resemble the observed) could be justified in a way that was both rational and non-circular. First, the principle of induction does not express any ‘relation of ideas’. The contradiction of a relation of ideas is inconceivable (or nonsensical), whereas the contrary of any induction, no matter how well-supported,

is always conceivable: ‘That the sun will not rise tomorrow is no less intelligible a proposition, and implies no more contradiction, than the affirmation, that it will rise’ (EHU 4.2 / 25–6). However, neither is the principle of induction a ‘matter of fact’.1 All inferences from experience must presuppose the principle of induction; hence that principle is too fundamental to be justified by appeal to experience: ‘All inferences from experience suppose, as their foundation, that the future will resemble the past’ (EHU 4.21 / 37). Likewise, attempts to justify induction by reference to the uniformity of nature face the insuperable obstacle that any belief in such uniformity can itself only be justified by induction. So while the principle of induction and the uniformity of nature are certainly not meaningless, neither are they susceptible to rational, non-circular justification. However, Hume did not think all inductions were on a par; still less did he advocate facing the future with inductive paralysis. Creatures like us possess a kind of mental inertia that preserves our mental states in being.2 Hence we instinctively project those regularities we observe. Furthermore, our expectation that such regularities will continue should be proportional to the evidence in their favour: ‘The creature expects from the present object the same consequences, which it has always

227

9780826443595_Ch11_Final_txt_print.indd 227

2/9/2012 9:36:14 PM

HUME ON MIRACLES found in its observation to result from similar objects’ (EHU 9.4 / 106). Thus, Hume’s pragmatic solution to inductive scepticism urges the maintenance of due proportion between degree of belief and evidence. However, even ideally well-confirmed inductions may fail. While proportioning degree of belief to available evidence is the best we can do, nature need not conform to our inductions or vindicate our predictions. Inductive failure is always conceivable and even a small degree of non-uniformity makes us hesitant in projecting regularities. Therefore, there is a sharp distinction between unbroken regularities and those that have admitted of exceptions. The projection of very widely observed, uniform regularities is the most powerful epistemic norm there is. So Hume is not recommending mere sceptical rejection of inductive practices, but instead arguing that unaided reason is not the primary regulator of our inductions. Equally importantly, he prescribes how induction should be regulated: while widely supported, hitherto unbroken uniformities can always fail; we should have the strongest degree of personal conviction that they will continue – interruptions are always possible but, prior to experience, they should be incredible. Inductively wise creatures obey their projective instincts. Crucially, Hume had a three-fold distinction between demonstrations, proofs and probabilities. A demonstration is essentially a deductive argument, one appealing to relations between ideas. (So Pythagoras’ Theorem would be amenable to demonstration.) Proofs and probabilities are inductive arguments, differing only in their evidential support. A proof is the strongest possible argument from experience – i.e., the projection of a well-confirmed, hitherto exceptionless regularity.3 (We have a proof that the

sun will rise tomorrow – no exceptions are recorded and many positive instances are.) However, a probability involves projecting a well-supported but not exceptionless regularity. (Thus, if your car has started on the first try some 99 previous mornings out of 100, you should have a strong probability for its starting first try tomorrow, but not a proof.) So, odd though it may sound to modern ears, Humean proofs are defeasible and may be opposed by greater proofs.4 When proofs conflict, we should incline our beliefs to that proof with the better evidential support both quantitatively and qualitatively. Key to Hume’s argument is that a miracle is characterized as ‘a violation of the laws of nature’ (EHU 10.12 / 114) – i.e. an exception to one of the best-confirmed regularities we possess. At times he seems to take this to be a complete characterization.5 There is a popular but misleading interpretation of ‘Of Miracles’ which we shall call the strong reading. On this view, Hume argues a priori (i.e. by appealing to general principles) that testimony can never rationally ground belief in miracles.6 That is, however strong the testimony, it cannot provide compelling evidence for a miracle because any epistemic force the testimony has will inevitably be outweighed by the evidence one has against the possibility of miracles. Taken in isolation, parts of ‘Of Miracles’ may favour this reading by suggesting there will always be more reason to suppose testimony to a miracle should be rejected than that a miracle has occurred. Consider this passage, for example: ‘no testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous, than the fact, which it endeavours to establish’ (EHU 10.13 / 115–6). On the strong reading, Hume claims that if we hear testimony to an event contrary to natural law,

228

9780826443595_Ch11_Final_txt_print.indd 228

2/9/2012 9:36:15 PM

HUME ON MIRACLES then rationality obliges us to reject the testimony as unreliable. So if someone tells you Jesus miraculously turned water into wine, you necessarily have more reason to reject the testimony than to suppose this miracle occurred, and hence this testimony provides you with no rational basis on which to believe this miracle took place. However, the strong reading is multiply suspect. For one thing, it fails to account for passages where Hume explicitly allows that testimony can make belief in a miracle rationally compelling. Another popular misreading trivializes Hume’s case by making him claim that the very concept of a miracle makes it a contradictory or impossible event. After all, if laws of nature are defined as inviolable, then it trivially follows that transgressions of laws of nature are impossible, and hence miracles must be impossible too. No wonder, then, that it would not be rational to believe in miracles on a testimonial (or any other) basis! By ruling out miracles by definition, this strategy would give opponents of miracles a suspiciously quick victory. However, this strategy is demonstrably not Hume’s. Hume never argues that a miracle is (conceptually or otherwise) impossible – indeed he explicitly allows not only that miracles may occur but that they can (at least in theory) be made the object of compelling testimony. Even the strong reading of Hume does not reject the very idea of miracles, but appeals only to the balance of evidence against their occurrence, given one’s overwhelming grounds in support of the uniformity of nature and the slender opposing grounds in favour of the miracle supplied by testimony. Consider this passage: Hume is not arguing that the wise reject testimony for miracles because they recognize that miracles are impossible by

definition. Nor is he claiming that noone could ever observe a miracle. He is not missing the point by defining ‘miracle’ in such a way that any event that actually occurs or is observed, no matter how bizarre, would fail to be a miracle.7 It is important to recognize that what miracles oppose is, for Hume, not an inviolable law of nature, but rather the widely observed regularity of a natural uniformity. Thus, miracles on this conception are certainly not impossible, but they are problematic objects of inference and things which we should be at the very least reluctant to believe in.8 A useful summary is this: Hume is speaking in descriptive mode, in which a law of nature is a type of phenomenon that has so often been found to occur that, our innate expectations being of the kind they are, we cannot help but assign it a maximal or near-maximal probability of always being repeated in the same circumstances. There is, however, no implication that we are correct to do so.9 As an example of correct proportionality between belief and evidence, consider Hume’s (EHU 10.10 / 113–14) Indian prince. Having lived all his life in warm climes and never observed water freeze, the prince receives testimony that water can turn into ice. The prince rejects this testimony because ice runs so counter to his own observations and thus to what he believes are the prevailing natural laws for water. Ice to the Indian prince seems an impossible extrapolation from water’s observed behaviour. Hume argues that the prince has greater reason to reject the testimony than to alter his beliefs in the light of this testimony, and hence he has ‘reasoned justly’ in rejecting such testimony (EHU 10.10

229

9780826443595_Ch11_Final_txt_print.indd 229

2/9/2012 9:36:15 PM

HUME ON MIRACLES / 114). Of course, the prince would not be correct in clinging dogmatically to his rejection of ice if his evidential basis changed – for example, if he were transported to Muscovy and saw water become solid with his own eyes. However, when the prince’s sole evidential support for the existence of ice is testimonial, Hume clearly thinks rejecting such testimony is more rational for the prince than acceptance. When evaluating testimony, rationality does not so much prescribe which beliefs are to be accepted, but rather how to change degrees of belief in the face of new evidence. While the prince reasoned to a false conclusion in rejecting the existence of ice, he nonetheless updated his beliefs as rationally as his evidence permits: A wise man, therefore, proportions his belief to the evidence. In such conclusions as are found on an infallible experience, he expects the event with the last degree of assurance, and regards his past experience as a full proof of the future existence of that event. (EHU 10.4 / 110) Note that the prince is not being offered testimony to a miracle, since the testimony claims only that water behaves differently in colder climes.10 In contrast, when someone testifies that a miracle has occurred – for example, that someone has walked on water – they explicitly claim that an event contrary to natural law has taken place, and this makes it even harder to accept the testimony: On the one side we have wide and unproblematic testimony to the effect that when people step into water they do not remain on its surface. On the other side we have isolated reports of people walking on the surface of water. Given

testimony of the first kind, how are we to evaluate testimony of the second sort? The testimony of the first sort does not show that the testimony of the second sort is false; it does, however, create a strong presumption – unless countered, a decisively strong presumption – in favour of its falsehood.11

In opposing natural law, testimony to a miracle incurs a greater testimonial opposition than would be the case if the event were merely unusual. The expectation that natural laws hold and will continue to hold, although by no means infallible, is nonetheless among the most powerful rational expectations there are, and any testimony that aims to overturn such expectations faces an uphill task. Barring remarkably powerful testimonial evidence to the contrary, the rational expectation must always be that natural laws are obeyed and that our reasoning from experience must proceed analogically. For the prince to believe in the existence of ice on the basis of testimony alone would offend against correct analogical reasoning: ‘The operations of cold upon water are not gradual, according to the degrees of cold; but whenever it comes to the freezing point, the water passes in a moment, from the utmost liquidity to perfect hardness’ (EHU 10.10n22 / 114n). It is not merely that the prince has not heard testimony that water does not turn to ice in cold climates; rather the process attested to implies a failure of natural analogical reasoning. Presumably the ice-testimony the prince receives is not very detailed – closer to ‘travellers’ tales’ in effect. But what if a miracle is extremely well attested to? Suppose hundreds of putative, mutually consistent witnesses testify. Would it not then be entirely unreasonable to dismiss the possibility that

230

9780826443595_Ch11_Final_txt_print.indd 230

2/9/2012 9:36:16 PM

HUME ON MIRACLES the miracle occurred? Hume is fully aware of this problem: ‘suppose, that all historians who treat of England, should agree, that, on the first of January 1600, Queen Elizabeth died . . . and that, after being interred a month, she again appeared, resumed her throne, and governed England for three years . . .’ (EHU 10.37 / 128). On the strong reading, Hume’s way of responding to such cases is not to weaken his stance in the context of the clear wealth of testimonial support in play. Instead, the ‘strong reader’ would have Hume argue that the extent of testimonial evidence on offer in these cases means only that we should seek out the natural causes which gave rise to this extraordinary event and thereby determine why so many people are under the false impression that a genuine miracle took place. This would be akin to the prince moving to ‘Muscovy during the winter’ (EHU 10.10n22 / 114n) in order to see if water does indeed turn to ice as has been alleged. What the prince will discover then is that there is nothing miraculous about this event at all, and hence no rational bar on this score for supposing that the target event occurs. The additional strength of testimonial support for the putatively miraculous event thus only has a bearing on the extent to which we are rationally obliged to discover a natural cause for that event, as opposed to simply dismissing the testimony tout court. Crucially, additional testimonial support cannot on the strong reading of Hume’s view ever suffice to give us reason to think that a miracle has occurred. While the strong reading has been (and remains) popular among Hume critics, it faces multiple problems. The fundamental problem with the strong reading is simply that it relies on partial readings of key passages in ‘Of Miracles’. Hume explicitly countenances

at least the possibility of a rationally formed, testimony-based, belief in the occurrence of a miracle: I beg the limitations here made may be remarked, when I say, that a miracle can never be proved, so as to be the foundation of a system of religion. For I own, that otherwise, there may possibly be miracles, or violations of the usual course of nature, of such a kind as to admit of proof from human testimony; though, perhaps, it will be impossible to find any such in all the records of history. (EHU 10.36 / 127) So clearly Hume believed (i) that miracles can occur and, crucially, (ii) that they can (at least in theory) be the subject of rationally compelling testimony. Hume is concerned particularly with those miraculous events that are capable of acting as support for a system of religion – that is, ‘testimonial’ miracles, miracles that testify to the divine mission, inspiration or guidance of a miracle-worker. So what is at issue is whether miracle-testimony could ever be good enough to make a miracle the foundation of a religious hypothesis (a ‘foundational miracle’). Hume does not merely mention the abstract possibility of acceptance-compelling miracle-testimony, he goes on to offer a detailed proposal for what an evidentially compelling miracle might look like: Thus, suppose, all authors, in all languages, agree, that, from the first of January 1600, there was a total darkness over the whole earth for eight days: suppose that the tradition of this extraordinary event is still strong and lively among the people: that all travellers, who return from foreign countries, bring us accounts of the same tradition, without the least variation or contradiction: it is evident,

231

9780826443595_Ch11_Final_txt_print.indd 231

2/9/2012 9:36:16 PM

HUME ON MIRACLES that our present philosophers, instead of doubting the fact, ought to receive it as certain, and ought to search for the causes whence it might be derived. The decay, corruption, and dissolution of nature, is an event rendered probable by so many analogies, that any phenomenon, which seems to have a tendency towards that catastrophe, comes within the reach of human testimony, if that testimony be very extensive and uniform. (EHU 10.36 / 127–8) Thus while Hume clearly thinks rational compelling testimony to miracles may be hard to come by, nonetheless it is clearly not impossible, contra the strong reading. For example, testimony can be compelling if it is ‘very extensive and uniform’, as it is in the ‘eight days of darkness’ case in contrast to the relatively parochial ‘Queen Elizabeth’ case. (The latter miracle would be too spatiotemporally localized to lend itself to the sort of testimony which would render it worthy of belief.) Likewise, the testimony can be compelling if it comes from widely separated witnesses who otherwise have little in common and who are not partisans for a particular explanation of the occurrence testified to, as opposed to (for example) historians with a particular wish to glorify Queen Elizabeth or to testify to divine endorsement of her policies. Even if Elizabeth’s resurrection was universally endorsed by historians, the balance of proof would still favour scepticism. Thus, the strong reading fails. Provided the above restrictions are met, testimony can be sufficient to support a rationally held belief in miracles.12 Nonetheless, while interruptions to the ordinary course of nature are not impossible, our epistemic nature being what it is, such interruptions will be (and ought to be) incredible, all else being equal. In most conceivable, and maybe all recorded, cases,

we would do better to reject the testimony than believe a miracle has occurred. Hume thus believes that testimony to a miracle will almost certainly not be of sufficient quality and quantity to make belief rationally compelling. Furthermore, even granted belief in the miracle itself is compelling, it is exceedingly unlikely, if not impossible, for the testimony to be so powerful as to make it rational to believe that the miracle in question testifies uniquely to the truth of a particular religious hypothesis. Call this the weak reading.13 While the strong reading does seem to have some textual support in its favour, it fails to pay due attention to two key, and related, features of Hume’s approach. The first is the essentially empirical nature of Hume’s discussion. Hume is at least as much concerned to offer an adequate description of our epistemic and evidential practices as he is to legislate for those practices. The second concerns the need to distinguish sharply between Hume’s treatment of miracles per se, and his treatment of specifically religious miracles – viz., a miracle involving a divine intervention into the natural order, such as might serve as the ‘foundation of a system of religion’ (EHU 10.35 / 127). This last restriction – that is, to miracle-testimony aimed at supporting a particular religious hypothesis – is very important and, contrary to Hume’s own cautions, one often ignored by critics. It is tempting to regard Hume as aiming to offer some sort of a priori limitation on what it is reasonable to believe regarding miracles. After all, a good deal of what he says, particularly in Part 1, is seeming general reflection on the nature of the epistemology of testimony as it applies to miracles. Crucially, however, it would be utterly contrary to the general spirit of Hume’s philosophy to establish a matter of fact in any a priori fashion, and hence we should be immediately suspicious of this way

232

9780826443595_Ch11_Final_txt_print.indd 232

2/9/2012 9:36:16 PM

HUME ON MIRACLES of reading him. The example of the eight days of darkness bears this suspicion out, since Hume is clearly not ruling out the possibility of a testimony-based, rational belief in the existence of a miracle tout court. Instead, in offering general remarks on the nature of testimony, Hume argues that the rational bar is set extremely high for testimony-based beliefs about miracles. He then appeals to empirical facts about cases where miracles are testified to and argues that it would be exceedingly hard for any testimony-based belief in miracles to clear this bar. Hume does not forbid miracles a priori or think such occurrences could never under any conceivable circumstances be made the object of rationally compelling testimony. Rather, reflection on correct epistemic practice reveals that the hurdle facing miracle-testimony is so high that in practice there is a negligible likelihood that real miracle-testimony can be found which is compelling. This brings us to the second point, for Hume is quite clear that there are empirical grounds for treating our testimony-based beliefs in the existence of specifically religious miracles as even more epistemically dubious than testimony-based beliefs in the existence of non-religious miracles. It is one thing to come to believe, on the extensive testimonial basis described, that the Earth was plunged into darkness for eight days, contrary to one’s current understanding of the natural order. It is another thing to believe that this event was a religious miracle, since this will impose an additional epistemic burden. As Robert Fogelin puts it: With respect to miracles, Hume’s strategy is to use the canons of causal reasoning to evaluate testimony brought forward in their behalf. Because, for him, no matter of fact can be established a priori, it

remains an open, though remote, possibility that testimony could establish the occurrence of a miracle. That is the point of the discussion of the eight days of darkness. With respect to miracles intended to serve as the foundation of a religion, the situation, according to Hume, is factually different. When we examine the testimony brought forward on their behalf, we see that it has uniformly failed to meet appropriate standards of acceptability.14 When dealing with testimony-based beliefs regarding (mere) miracles versus those regarding specifically religious miracles, the issue is fundamentally an empirical one related to whether there is sufficient epistemic support for the testimony-based belief in question to clear the bar Hume lays down for it. Hume’s claim, however, is that any testimony-based belief in a miracle which is also saddled with the burden of a religious explanation would require an even greater degree of epistemic support to clear this bar, and so faces an even stiffer challenge (one which it is hard, if not impossible, to see any actual belief being able to satisfy in practice). Another difficulty for the strong reading of Hume’s thesis is that it implies that the business of ‘Of Miracles’ is effectively completed by the end of Part 1. In Part 1, Hume proposes a criterion for what counts as a miracle and points out that testimony to a miracle is of its very nature testimony to an occurrence that violates one of our bestsupported regularities. However, while Part 1 argues that the evidential benchmark for miracles is set very high, it does not establish that this benchmark is impossibly high or unattainable in practice. Part 1 at least leaves open the theoretical possibility of compelling testimony to the occurrence of miracles. Thus, Part 1 draws out a conflict

233

9780826443595_Ch11_Final_txt_print.indd 233

2/9/2012 9:36:16 PM

HUME ON MIRACLES between two intuitions: one that testimony is ordinarily worthy of high (nay decisive) epistemic credit and another that scepticism about the continuance of laws of nature is hard to sustain. These two intuitions ordinarily command our assent without coming into conflict. Thus, the problem posed by miracle-testimony is that of how to adjudicate between two ordinarily compelling intuitions forced into stark confrontation. Fogelin argues that Part 1 aims only to emphasize this tension and does not pronounce on how it should be resolved. Part 2 then develops the case against miracles by invoking empirical considerations about observed testimony to extraordinary events. In Part 2, Hume offers empirical grounds why testimony to miracles is likely to be especially untrustworthy. For example, human beings want to believe that miracles occur because this leads to ‘agreeable emotion[s]’ (EHU 10.16 / 117), and this may explain why history is littered with stories of ‘forged miracles’: The many instances of forged miracles, and prophesies and supernatural events, which, in all ages, have either been detected by contrary evidence, or which detect themselves by their absurdity, prove sufficiently the strong propensity of mankind to the extraordinary and the marvellous, and ought reasonably to beget a suspicion against all relations of this kind. (EHU 10.19 / 118) Hume also argues that ‘civilised’ societies move away from supernatural and/or miraculous explanations: It forms a strong presumption against all supernatural and miraculous relations, that they are observed chiefly to abound among ignorant and barbarous

nations; or if a civilised people has ever given admission to any of them, that people will be found to have received them from ignorant and barbarous ancestors, who transmitted them with that inviolable sanction and authority, which always attend received opinions. (EHU 10.20 / 119)

Hume is thus arguing that the development of civilization – and thus the rejection of ‘ignorant and barbarous’ traditions – involves curtailing the natural impulse towards treating events as miracles. Hume then allows himself a little sarcasm: ‘It is strange, a judicious reader is apt to say, . . . that such prodigious events never happen in our days. But it is nothing strange, I hope, that men should lie in all ages.’ (EHU 10.21 / 119–20) Thus far, of course, these empirical grounds against accepting testimony to miracles will apply to any miracle. Hume clearly thinks, however, that empirical concerns about testimony are especially acute when it comes to specifically religious miracles. Of their very nature, religious miracles are often promoted to favour a particular cause, and thus we can reasonably attribute substantial self-interest to those putting forward the initial testimony. But as Hume (EHU 10.21–3 / 120–1) makes clear, where self-interest is involved in testifying to miracles, one should be wary of that testimony. Moreover, because miracles are often associated with particular religions (or sects) it is unsurprising that there is in fact a great deal of conflict in the kind of miracle reports that are found in the great religions (EHU 10.24 / 121–2). Contrariety to natural laws already places a stiff burden of proof on miracle enthusiasts, but applying our knowledge of human psychology to such testimony makes the evidential situation still worse. We know people tend to embellish stories likely

234

9780826443595_Ch11_Final_txt_print.indd 234

2/9/2012 9:36:16 PM

HUME ON MIRACLES to incite wonder. Miracle-testimony is usually at several removes from witnesses – that is, it testifies to miracles far away and/or long ago.15 Furthermore, the diversity of miraclereports from different religious traditions reduces the force of miracle-testimony in any particular case. Finally, there are natural explanations for why people want to believe in specifically religious miracles: The wise lend a very academic faith to every report which favours the passion of the reporter; whether it magnifies his country, his family, or himself, or in any other way strikes in with his natural inclinations and propensities. But what greater temptation than to appear as a missionary, a prophet, an ambassador from heaven? Who would not encounter many dangers and difficulties, in order to attain so sublime a character? Or if, by the help of vanity and a heated imagination, a man has first made a convert of himself, and entered seriously into the delusion; who ever scruples to make use of pious frauds, in support of so holy and meritorious a cause? (EHU 10.29 / 125) Religious believers face temptations that testifiers to merely secular extraordinary occurrences do not face. To appear as the medium of divine revelation might tempt otherwise utterly scrupulous witnesses to embroider, distort, omit or invent. The point here is two-fold. On the one hand, nothing would exercise the passions as much as being able to present oneself as a witness to a religious miracle. On the other hand, if one is already inflamed with religious passion, one has reason to accept testimony in support of a religious miracle, regardless of any epistemic basis for supposing this testimony reliable. Both considerations offer specific empirical grounds against trusting

testimony concerning the occurrence of religious miracles. There is another reason why Hume thinks that any actual miracle-testimony thus far recorded cannot support religious hypotheses, and this is that miracle-traditions are live parts of the traditions of a host of competing and mutually exclusive religious faiths: ‘all the prodigies of different religions are to be regarded as contrary facts, and the evidences of these prodigies, whether weak or strong, as opposite to each other’ (EHU 10.24 / 122). Inevitably, different faiths will make different claims about the doctrines to which their miracles testify. Of course, it is conceivable that the divinity might depute miracle-working power to several miracle-workers. (After all, Acts 8: 18–20 never describes Simon Magus, the sorcerer who tried to buy spiritual power from Peter, as attempting the logically impossible.) However, the possibility of disparate miracle-workers is something Hume does not need to rule out a priori. Hume might retort that several miracle-workers performing miracles in support of different faiths is of itself enough to undercut the claim that performing a miracle uniquely testifies to one particular faith. It is in this sense that Hume clearly regards testimony to miracles in support of different faiths as mutually destructive. However, it is not clear that reports of miracles from different traditions are logically in conflict, even if the religions thereby supported make conflicting claims. So despite some textual support for the strong reading, the evidence actually supports the weaker reading. Crucially, however, the weak reading is just as challenging to the idea that one should have a rationally held, testimony-based belief in the occurrence of a miracle, much less a religious miracle. For even if it is possible for such a belief to be rationally held, the epistemic bar that must

235

9780826443595_Ch11_Final_txt_print.indd 235

2/9/2012 9:36:17 PM

HUME ON MIRACLES be cleared in order for this to be the case is so high that it is hard to see how, practically speaking, any belief could clear it nor that any belief actually has cleared it.16 In what follows we will explore some further issues of relevance in this regard, beginning with the supposed import Hume’s remarks on the epistemology of testimonybased belief in miracles have been thought to have for the epistemology of testimony in general.

2. HUME AND THE EPISTEMOLOGY OF TESTIMONY In the contemporary epistemological literature Hume is often characterized as holding a rather radical view about the epistemology of testimony. According to this interpretation – popularized by C.A.J. Coady17 – Hume’s remarks about testimonybased belief concerning miracles reveal a general reductionist view about the epistemology of testimony-based belief. There are a number of competing ways of drawing the reductionism/anti-reductionism distinction in the contemporary epistemological literature on testimony, but very roughly reductionism holds that the epistemic standing of a testimony-based belief must ultimately be completely traced back to, and hence in this sense reduced to, non-testimonial sources, while anti-reductionism denies this claim.18 In practice, reductionism means that agents can never simply rely on testimony if they want to have beliefs which enjoy appropriate epistemic support. Instead, they must always seek independent grounds in favour of the target belief. So, for example, in order to form an epistemically sound belief on the basis of testimony, one requires such independent

grounds as, for instance, previous first-hand experience of the reliability of this informant on this subject matter and collateral information regarding the plausibility of the proposition testified to. There is a good rationale that can be offered for reductionism. For one thing, it is widely held that testimony is only a transmissive, and thus not a generative, source of knowledge. That is, testimony can only at best transfer knowledge that has already been acquired, but cannot be used to acquire new knowledge (testimony is held to be similar to memory on this score).19 If that is right, then it seems that the epistemic status of a testimony-based belief must be ultimately derived from a non-testimonial source. Moreover, as Elizabeth Fricker has argued, to reject reductionism is, it seems, to put an epistemic premium on gullibility.20 Why should the mere fact that someone testifies that p give you any epistemic basis, however modest, for believing that p? There also seems good reason for thinking that Hume’s argument against the rationality of testimony-based beliefs in the existence of miracles draws upon a reductionist account of the epistemology of testimony. Consider, for example, the following oft-cited passage: ‘our assurance in [testimony] is derived from no other principle than our observation of the veracity of human testimony, and of the usual conformity of facts to the reports of witnesses’ (EHU 10.5 / 111). Here Hume’s sympathy with reductionism appears perfectly clear: we gain epistemic support from testimony only when we have an independent basis for the belief so formed, such as from our first-hand observations about the ‘veracity of human testimony’. The key problem with reductionism, however, is that it appears to entail that not only are we unable to gain epistemically well-grounded

236

9780826443595_Ch11_Final_txt_print.indd 236

2/9/2012 9:36:17 PM

HUME ON MIRACLES beliefs about the occurrence of miracles, but also about much else besides. After all, a great deal of our beliefs were acquired through testimony,21 and yet in a large number of cases not only do we have no practical independent (i.e. non-testimonial) means of verifying the target proposition but we also cannot even remember what the particular testimonial source of our belief was. Accordingly, it seems that reductionism entails a fairly widespread scepticism about the epistemology of testimonial belief. Right now, for example, I believe – indeed, I think I know – that the Orinoco River flows through Venezuela, but although I know this belief must have been acquired via testimony, I have not the slightest recollection of the circumstances in which it was acquired. As such, I can hardly have any independent epistemic basis for trusting the testimonial source for this belief. Moreover, although I have various means at my disposal for checking this belief, unless I actually take the trouble to travel to South America to determine the matter in person, I will almost certainly need to depend on further instances of testimony, such as an atlas or the testimony of a colleague, in carrying out these checks. Hence, it is hard to see how I could rationally hold this belief by reductionist lights. But what applies to this belief will also apply to many other beliefs that I hold. The upshot is that if reductionism is true then we have far less epistemically well-grounded, testimony-based beliefs than we tend to suppose. In itself, of course, this is a fairly indecisive strike against reductionism. After all, perhaps we do know a lot less than we think we do. Still, reductionism is a contentious thesis in the literature on the epistemology of testimony, with the dominant camp being by far anti-reductionism.22 Accordingly, insofar as we treat Hume’s argument against the

rationality of testimony-based beliefs in the existence of miracles as trading on this thesis then that might be thought to be a pretty serious count against it. Fortunately for Hume, however, it is clear on closer inspection that not only does his argument against the rationality of testimony-based belief in miracles not trade on reductionism, but it is also questionable whether he assents to this general thesis about the epistemology of testimony anyway. We will take these points in turn. The first point can be established by noting that the anti-reductionist could quite consistently endorse Hume’s argument against the rationality of a testimony-based belief in the existence of miracles (on either the strong or the weak reading). After all, what is key to anti-reductionism is only the claim that in epistemically suitable circumstances one can gain an epistemically wellfounded testimony-based belief even while lacking the full, independent, epistemic support demanded by reductionism. That is, while the reductionist demands that independent epistemic support is always required for a testimony-based belief to be rationally held, the anti-reductionist demurs and argues that at least sometimes – that is, in epistemically propitious circumstances – such a belief can be rationally held even in the absence of independent epistemic support. It is thus open to the anti-reductionist to claim that the testimonial beliefs at issue when it comes to miracles are not formed in the required conditions and hence cannot benefit from the greater epistemic permissiveness of the antireductionist thesis.23 Indeed, there is in fact every reason for thinking that an anti-reductionist would be sensible to take this line when it comes to testimony-based beliefs in the existence of miracles. For given that miracles are by their

237

9780826443595_Ch11_Final_txt_print.indd 237

2/9/2012 9:36:17 PM

HUME ON MIRACLES nature the kind of event that does not normally occur – indeed, as we noted above, on a certain conception of miracles (though not the one that Hume had in mind) they may well be the kind of event that cannot occur – it follows that there is always a standing reason to doubt the veracity of testimony regarding miracles. The kind of testimonial cases that the anti-reductionist wants to protect from the zealous epistemic strictures of reductionism, however, are precisely those where there is no standing basis for doubt. It is one thing to argue that in good epistemic conditions one can gain an adequate epistemic basis for one’s testimony-based belief even while lacking any independent support for that belief, and quite another to claim that an adequate epistemic basis can be had without independent support where the epistemic conditions are problematic in some way. The case of testimony regarding the occurrence of a miracle clearly falls into the second, problematic, category, on account of there being in such a case a standing doubt about the veracity of the testimony in question. In such instances the anti-reductionist might well demand – indeed, would be wise to demand – that the agents concerned should seek independent grounds for their belief. Moreover, note that making the reductionist demand for independent epistemic support in the specific case of testimony regarding miracles would not settle the matter of whether a belief in this regard could be rationally held. For as the ‘eight days of darkness’ example illustrates, once the wealth of testimonial evidence becomes particularly extensive, it is far from clear that one does lack an independent basis for taking the testimony at face-value. After all, one can determine on a priori grounds that where there is a uniform convergence in the testimony regarding a significant event that is

offered by all people of all civilized nations, then that is a good epistemic basis on which also to form a belief in this proposition, even if one cannot verify the truth of that proposition oneself. Accordingly, even if we were to apply reductionist standards to testimonybased belief in the existence of miracles, it is still not obvious that this would thereby necessarily prevent such a belief from being rationally held. This brings us to our second point, which is that it is questionable whether Hume would in any case assent to the general reductionist thesis regarding the epistemology of testimony. The foregoing should make it clear why. For if Hume’s argument against the rationality of testimony-based belief in the existence of miracles is compatible with anti-reductionism, then clearly we have no grounds for concluding, on the basis that he appears to apply reductionist standards in this particular case, that he is in general in favour of reductionism. Indeed, we could just as well conclude on this basis that he is an anti-reductionist.

3. BAYESIAN VERSIONS OF HUME’S ARGUMENT It remains disputed whether Hume’s argument embodies correct probabilistic reasoning and whether it can be reconstructed in Bayesian terms. Responses here can be sharply polarized.24 Bayesianism takes its name from Hume’s near-contemporary, the Reverend Thomas Bayes (1702–61), whose ideas about probability were first published posthumously in 1763 by Hume’s correspondent Richard Price (1723–91).25 Bayes’ Theorem aims to quantify how degrees of belief in a hypothesis should vary in the light

238

9780826443595_Ch11_Final_txt_print.indd 238

2/9/2012 9:36:17 PM

HUME ON MIRACLES of new evidence (e.g. that better-supported theories should receive greater belief). In its simplest form, Bayes’ Theorem says that the probability a hypothesis receives conditional upon evidence equals the likelihood of the evidence if the hypothesis is true multiplied by the hypothesis’s prior probability and divided by the evidence’s prior probability. More formally:

So, we want to determine the posterior probability testimony (‘t’) confers on a miracle (‘M’) – i.e. Pr(M࿽t). Probabilities range between 0 and 1, 0 corresponding to certain falsehood and 1 to certain truth. One popular misreading sets Hume’s prior probability for a miracle – i.e. Pr(M) – at 0, from which normal Bayesian conditionalization inevitably dictates that Pr(M࿽t) = 0.26 However, setting Pr(M) = 0 treats Hume’s proofs as indefeasible and conflicts with the ‘eight days of darkness’ example. Setting Pr(M) = 0 foists onto Hume principles he did not hold. Consider this quotation, for example: To understand the structure of Hume’s argument, it is helpful to try to specify the form that Hume thinks inductive reasoning follows. As a starting point, recall Reichenbach’s straight rule of induction: if n As have been examined and m have been found to be Bs, then the probability that the next A examined will be B is m/n. Corollary: If m = n, then the probability that the next A will be a B is 1. . . . A (Hume) miracle is a violation of a presumptive law of nature. By Hume’s straight rule of induction, experience confers a probability of 1 on a presumptive law. Hence the probability of a miracle is flatly zero. Very simple. And very crude.27

And very hard to square with Hume’s text.28 Hume cannot accord extremal (0 or 1) probabilities to any empirical outcome and the ‘straight rule’ is not Humean.29 Although Hume talks of testimony ‘establishing’ a miracle, this does not mean testimony must support a miracle beyond all (reasonable) doubt – i.e. such that Pr(M࿽t) = 1. Rather, the occurrence of M is confirmed if t makes it more probable than not that M occurred – i.e. if the ‘balance of probabilities’ favours M rather than ¬M. Therefore, M is ‘established’ by testimony if Pr(M࿽t) > 0.5. In turn, Pr(M࿽t) > 0.5 if Pr(t & ¬M) < Pr(M) Pr(t࿽M). We can safely assume Pr(t࿽M) = 1 – i.e. had M occurred ‘then testimony to that effect would certainly have been forthcoming’.30 If M is to support a particular religious hypothesis, then presumably testimony to M must be forthcoming, given plausible assumptions about the raison d’être of miracles. Therefore, following Peter Millican, Hume’s rule is:31 Pr(M࿽t) > 0.5 → Pr(M࿽t) > Pr(~M࿽t)32 ‘Of Miracles’ requires two distinct arguments: first, that the prior probability of a miracle will be low, and second, that testimony to the occurrence of a miracle is likely to be forthcoming even if no miracle occurred33 (conclusions Hume derives in Parts 1 and 2 respectively). Therefore, Pr(M) should be very low and Pr(t & ¬M) very high; hence it is very unlikely Pr(M࿽t) > 0.5. Hume’s argument would be incomplete had he argued only for low Pr(M) – he also needs reasons for high Pr(t࿽¬M). For example, if we have very low Pr(M) yet think it very unlikely t would be forthcoming if M had not happened, Pr(M࿽t) can still be as close to 1 as desired. Critics of Hume often object that testimony might come from so many independent

239

9780826443595_Ch11_Final_txt_print.indd 239

2/9/2012 9:36:18 PM

HUME ON MIRACLES observers that the possibility of collusion or honest mistake can be made vanishingly low and therefore Pr(M࿽t) raised vanishingly close to 1.34 However, this Bayesian point is neither damaging to Hume nor one Hume need query. As noted above, he held that miracles can be objects of compelling testimony. The crucial point is whether we think he thought testimony can never establish a foundational miracle or (a weaker conclusion) that testimony has never established a foundational miracle. Even if we allow that miracles might (theoretically) be made the subject of compelling testimony in a way that supports a particular religious hypothesis, we might then look at the historical credentials of (e.g.) biblical miracles. Imagine that all nations around the Red Sea recorded the miraculous destruction of Pharaoh’s armies, that all nations recorded that the sun had halted over Gibeon,35 or that Roman historians recorded that Christ resurrected Lazarus before Tiberius. In such cases, the evidential position would be far better quantitatively and qualitatively than in any actual biblical case. Whether or not the historical David Hume would have become a believer had such (counterfactual) testimony existed, a Humean could set the evidential benchmark at this level without in any way contradicting Hume’s principles. Believers in miracles might wonder why actual miracle-testimony is not better. Any finite amount of testimony could (theoretically) be improved upon. However, the question is not ‘Why does actual miracle-testimony admit any possibility of doubt?’, but rather ‘Why is actual miracle-testimony so impoverished?’. Remember that Hume’s main conclusion is not that miracles cannot be established by testimony but that they (almost certainly) cannot be so established as to support a system of religion. From responses to criticism, clearly Hume

regarded compelling miracle-testimony as a hypothetical possibility, yet far from an actuality: There is no contradiction in saying, that all the testimony which ever was really given for any miracle, or ever will be given, is a subject of derision; and yet forming a fiction or supposition of a testimony for a particular miracle, which might not only merit attention, but amount to a full proof of it. (LDH 1.349, 188: Letter to Hugh Blair, 1761)

4. CONCLUDING REMARKS The above aims to situate Hume’s account of testimony regarding miracles in the context of his views of correct inductive practice, to guard against some popular misunderstandings, to explore how far he can be regarded as a reductionist about testimony, and to offer a few remarks on the relevance of Bayesianism to his argument. The history of ‘Of Miracles’ suggests that certain misreadings of Hume are too tempting to stay buried for very long, and no doubt claims will continue to be made that his account of miracles contradicts his philosophy of induction, or that he achieves a cheap victory by defining miracles out of existence. However, these claims are proving harder to sustain as time passes. Besides its continuing relevance to philosophy of religion and the epistemology of testimony, ‘Of Miracles’ remains a key source for Hume’s positive account of induction and a useful counter to any views that would dismiss him as an unreconstructed sceptic or unreflective scoffer. He anticipates in important ways the increasing trend to naturalistic, even evolutionary, explanations of our cognitive strategies and their success.

240

9780826443595_Ch11_Final_txt_print.indd 240

2/9/2012 9:36:23 PM

HUME ON MIRACLES Although by no means immune to criticism, his ‘Of Miracles’ is in no danger of triviality or obsolescence.36

NOTES 1

2

3

4

While Hume’s ‘relations of ideas’ and ‘matters of fact’ resemble analytic and empirical truths respectively, the pairs of terms are not synonymous. See W. Ott, Causation and Laws of Nature in Early Modern Philosophy (Oxford: Oxford University Press, 2009), pp. 200–3. Hume sought to extend Newtonian mechanical description of physical phenomena into the science of the mind. Thus, continuity of mental habits may not merely resemble inertial persistence of motion, but may be another instance of the same phenomenon – see S. Buckle, ‘Marvels, Miracles, and Mundane Order’, Australasian Journal of Philosophy 79 (2001), pp. 1–31. As he puts it, proofs are ‘arguments from experience as leave no room for doubt or opposition’ (EHU 6n / 56n). ‘[I]t may well seem strange that Hume should use the word ‘proof’ for an argument which he acknowledges to be fallible, i.e. nondemonstrative. However, his training and later employment as a lawyer would make it much more natural for him to use the word in its original and legal sense rather than in its now familiar (to philosophers anyway) mathematical and logical sense. In a legal context, proof is a matter of degree and a probable argument, in Hume’s sense, can be thought of as yielding a partial proof’ (B. Gower, ‘Hume on Probability’, British Journal for the Philosophy of Science 42 (1991), pp. 1–19, p. 4n). In any case, Hume here arguably follows our ordinary language usage of ‘proof’: we would normally think it an entirely adequate ‘proof’ that there are at least three misprints on a page simply to point out three separate misprints (cf. G.E. Moore, ‘Proof of an External World’, Proceedings of the British Academy 25 (1939), pp. 273–300, p. 275). According to Buckle, ‘Marvels, Miracles, and Mundane Order’, p. 7, for Hume: ‘A proof is not a demonstration, not a deductively valid argument – and therefore not what we would call a proof’. Buckle delineates two

5

6

7

8

9

Humean kinds of ‘proofs’: (i) where no doubt is possible, such as G.E. Moore’s external world proof; and (ii) ‘the limit case of probable argument, since it is that conclusion drawn from experience where the uniformity of experience is unalloyed’ (ibid., p. 8). For more on Hume’s ‘proofs’, see R. Fogelin, A Defense of Hume on Miracles (Princeton: Princeton University Press, 2003), pp. 13–17. At one point Hume (EHU 10.12n23 / 115n) does suggest that the putative transgression of natural law must be brought about by a supernatural agent: ‘A miracle may accurately be defined, a transgression of a law of nature by a particular volition of the Deity, or by the interposition of some invisible agent.’ According to this definition, a transgression of natural law brought about in some arbitrary fashion would not count as a miracle. We set this complication aside henceforth, not least because Hume describes alleged miracles which involve no supernatural agency, appear perfectly secular and have no apparent volitional origin – e.g. Queen Elizabeth’s imaginary resurrection (EHU 10.37 / 128). See, for example, D. Coleman, ‘Hume, Miracles, and Lotteries’, Hume Studies 14 (1988), pp. 328–46 and R.J. Fogelin, ‘What Hume Really Said about Miracles’, Hume Studies 16 (1990), pp. 81–7. Note that in more recent work Fogelin (A Defense of Hume on Miracles) rejects this reading. D. Garrett, Cognition and Commitment in Hume’s Philosophy (Oxford: Oxford University Press, 1997), p. 152. For more on the notion of a miracle, see R.F. Holland, ‘The Miraculous’, American Philosophical Quarterly 2 (1965), pp. 43–51; R. Young, ‘Miracles and Epistemology’, Religious Studies 8 (1972), pp. 115–26; M. Levine, ‘Miracles’, in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (online at: http://plato.stanford.edu/entries/miracles/); and D. Corner, ‘Miracles’, in B. Dowden and J. Feiser (eds), The Internet Encyclopedia of Philosophy (online at: www.iep.utm.edu/M/ miracles.htm). C. Howson, Hume’s Problem: Induction and the Justification of Belief (Oxford: Oxford University Press, 2000), pp. 241–2.

241

9780826443595_Ch11_Final_txt_print.indd 241

2/9/2012 9:36:23 PM

HUME ON MIRACLES 10

11 12

13

One might say that ice for the prince is marvellous (or extraordinary), and not miraculous – i.e. an exception to, not contradictory of, a law of nature. With ice: ‘Contrariety would lie in its going solid at a tropical temperature’, whereas ‘[r]esurrection involves more than an extension of the complex processes of nature: it involves something close to a reversal of those processes’ (M.A. Stewart, ‘Hume’s Historical View of Miracles’, in M.A. Stewart and J.P. Wright (eds), Hume and Hume’s Connexions (Edinburgh: Edinburgh University Press, 1994), pp. 171–200, pp. 194–5). Likewise, ‘[o]ther miracle stories, such as turning one substance into another, do not present the same shock to the system, the same undoing of the past, that resurrection stories do’ (ibid., p. 195). However, how far Hume can justifiably draw a miracles/ marvels distinction is controversial. For views con and pro see (for example) C.A.J. Coady, Testimony: A Philosophical Study (Oxford: Oxford University Press, 1992), chap. 10 and D. Coleman, ‘Baconian Probability and Hume’s Theory of Testimony’, Hume Studies 27 (2001), pp. 195–226, respectively. Fogelin, A Defense of Hume on Miracles, p. 20. Indeed, in the first of the two passages just cited Hume explicitly suggests that testimony can provide a ‘proof’ of the target event. To the modern ear it might sound as though Hume intended a deductive, and thus indefeasible, epistemic basis for the belief so formed, but as we have already noted above, this is not what Hume had in mind when he talked of ‘proofs’. The locus classicus for this reading is Fogelin, A Defense of Hume on Miracles. See also A. Flew, ‘Hume’s Check’, The Philosophical Quarterly 9 (1959), pp. 1–18 and Hume’s Philosophy of Belief: A Study of His First Inquiry, 2nd edn (Chicago, IL: St. Augustine’s Press, 1997). Textual evidence suggests Hume originally thought testimony to a miracle could never make belief in a particular religious hypothesis compelling but later adopted the ‘softer’ view that testimony to a miracle has never made belief in a particular religious hypothesis compelling. A crucial shift in wording comes in the third (1756) edition of the first Enquiry. In 1756 Hume says ‘no

14 15

16

17

18

19

20

testimony for any kind of miracle has ever amounted to a probability, much less to a proof’ (EHU 10.35 / 127), rather than using the earlier (1748 and 1750) formulation ‘no testimony for any kind of miracle can ever possibly amount to a probability’ (emphasis added). Fogelin, A Defense of Hume on Miracles, p. 62. In this regard, Hume implicitly refers to the argument found in John Craig’s Theologiae Christianae Principia Mathematica (1699), that ‘the longer the chain of testimony the less assurance it gives of the fact that a miracle had occurred’ (F. Wilson, ‘The Logic of Probabilities in Hume’s Argument against Miracles’, Hume Studies 15 (1989), pp. 255–75, p. 256). Craig calculated that any testimony-derived probability for Christianity would reach zero in c. 3150, and hence that this was a plausible date for the Second Coming. See S.M. Stigler, ‘John Craig and the Probability of History: From the Death of Christ to the Birth of Laplace’, Journal of the American Statistical Association 81 (1986), pp. 879–87. For a nice overview of Hume’s argument against the reasonableness of testimony-based beliefs regarding the existence of miracles, see P. Russell, ‘Hume on Religion’, in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (online at: http://plato.stanford.edu/entries/ hume-religion, sect. 6). See C.A.J. Coady, ‘Testimony and Observation’, American Philosophical Quarterly 10 (1973), pp. 149–55 and Testimony: A Philosophical Study. For more on the reductionism/anti-reductionism distinction in the epistemology of testimony, see D.H. Pritchard, ‘The Epistemology of Testimony’, Philosophical Issues 14 (2004), pp. 326–48 and J. Lackey, ‘Testimonial Knowledge’, in S. Bernecker and D.H. Pritchard (eds), The Routledge Companion to Epistemology (New York: Routledge, 2010). See J. Lackey, ‘Learning from Words’, Philosophy and Phenomenological Research 73 (2006), pp. 77–101 for a critique of this view of testimony. See E. Fricker, ‘The Epistemology of Testimony’, Proceedings of the Aristotelian Society, suppl. vol. 61 (1987), pp. 57–83;

242

9780826443595_Ch11_Final_txt_print.indd 242

2/9/2012 9:36:23 PM

HUME ON MIRACLES

21

22

23

24

25

26

‘Against Gullibility’, in B.K. Matilal and A. Chakrabarti (eds), Knowing from Words (Dordrecht: Kluwer Academic Publishers, 1994), pp. 125–61; ‘Telling and Trusting: Reductionism and Anti-Reductionism in the Epistemology of Testimony’, Mind 104, pp. 393–411. As Hume (EHU 10.5 / 111) says: ‘there is no species of reasoning more common, more useful, and even necessary to human life, than that which is derived from the testimony of men, and the reports of eye-witnesses and spectators.’ All else being equal, testimony has, and should have, exceptionally powerful epistemic force. For some of the main defences of antireductionism, see C.A.J. Coady, ‘Testimony and Observation’ and Testimony: A Philosophical Study; T. Burge, ‘Content Preservation’, The Philosophical Review 102 (1993), pp. 457–88; R. Foley, ‘Egoism in Epistemology’, in F. Schmitt (ed.), Socializing Epistemology: The Social Dimensions of Knowledge, (Lanham, MD: Rowman and Littlefield, 1994); and J. McDowell, ‘Knowledge by Hearsay’, in B.K. Matilal and A. Chakrabarti (eds), Knowing from Words (Dordrecht: Kluwer, 1994), pp. 125–61. For a recent discussion of how best to draw this contrast, see Pritchard, ‘The Epistemology of Testimony’. For an excellent and up-to-date overview of the literature on the epistemology of testimony, see Lackey, ‘Testimonial Knowledge’. Fogelin seems to have this point in mind when he claims that Hume’s reflections on the epistemology of testimony-based beliefs in the existence of miracles can be presented equally well within both a reductionist and an antireductionist account of the epistemology of testimony. See Fogelin, A Defense of Hume on Miracles, pp. 4–10, esp. p. 6n3). Witness the title of Earman (2000): Hume’s Abject Failure: The Argument Against Miracles (Oxford: Oxford University Press, 2000). T. Bayes; R. Price, ‘An Essay towards Solving a Problem in the Doctrine of Chances’, Philosophical Transactions of the Royal Society of London 53 (1763), pp. 370–418. J.H. Sobel, ‘On the Evidence of Testimony for Miracles: A Bayesian Interpretaion of David Hume’s Analysis’, Philosophical Quarterly

27 28

29

30 31

37 (1987), pp. 166–86, p. 176, suggests that Pr(M࿽t) = i /(i + Pr(~M & t)), where ‘i’ is a positive infinitesimal. However: Sobel treats our assurance in a law of nature as having probability infinitely close to one, and the corresponding violation as infinitely close to zero. I prefer to think of Hume’s notion of proof as being simply an argument with very high probability indeed. This allows there to be such a thing as a superior proof without resort to the unHumean notion of “infinitely close to”, and thus allows one to treat seriously Hume’s important example of the real possibility of being convinced that the earth was covered in darkness for eight days. (D. Owen, ‘Hume versus Price on Miracles and Prior Probabilities: Testimony and the Bayesian Calculation’, Philosophical Quarterly 37 (1987), pp. 187–202, p. 189n). Earman, Hume’s Abject Failure, pp. 22–3. For a countervailing view to Earman, see Fogelin, A Defense of Hume on Miracles. Cf. ‘Projection’: ‘when all the observed experiments lead to the same outcome O, there is probability 1 that an unobserved experiment leads to O’ (A. Mura, ‘Hume’s Inductive Logic’, Synthese 115 (1998), pp. 303–31, p. 311). Hume’s view of projection seems close to that of Carnap, who ‘did not view projection as an axiom of inductive logic, because in his view it is completely counterintuitive. Indeed, Carnap referred to it only as an illustration of the severe disadvantages of the straight rule’ (ibid.). Howson, Hume’s Problem, p. 244. P. Millican, ‘“Hume’s Theorem” concerning Miracles’, Philosophical Quarterly 43 (1993), pp. 489–95, p. 490. Note, however, that Millican expresses reservations about this reading (pp. 490–1, 495n8), and has recently developed an alternative ‘type’ interpretation in the context of a broader discussion of Hume’s essay: see P. Millican, ‘Twenty Questions about Hume’s “Of Miracles”’, in A. O’Hear (ed.), Philosophy and Religion (Cambridge: Cambridge University Press, 2011), pp. 151–92, sects 4–9. But as Millican points out (sect. 19), this alternative interpretation makes Hume’s rule strictly incorrect, though potentially revisable in a Humean spirit.

243

9780826443595_Ch11_Final_txt_print.indd 243

2/9/2012 9:36:23 PM

HUME ON MIRACLES 32

33

34

Sobel, ‘Hume’s Theorem on Testimony Sufficient to Establish a Miracle’, p. 232, summarizes Hume thus: Pr(t) > 0 & Pr(M࿽t) > ½ → Pr(M) > Pr(t & ~M) (substituting ‘M’ and ‘t’ for Millican’s and Sobel’s ‘A’ and ‘α’ respectively). A consequence of Bayes’s theorem is that Pr(M/t) = Pr(M & t)/Pr(t) – i.e. the posterior probability conferred on M by t equals the prior probability of the joint occurrence of M and t, divided by the prior probability of t. Howson, Hume’s Problem, p. 245, credits this insight to Charles Babbage (1791–1871). See the extracts from Babbage’s Ninth Bridgewater Treatise in Earman, Hume’s Abject Failure, pp. 203–12.

35

36

‘The sun stayed in the midst of heaven, and did not hasten to go down for about a whole day’ (Joshua 10: 13, Revised Standard Version). Although Hume does not say so explicitly, this planetary-scale miracle presumably stands to the ‘eight days of darkness’ rather as the resurrections of Lazarus and Jesus do to the resurrection of Elizabeth I. Many thanks to Peter Millican for discussions and advice. We are grateful to the editors of this volume for inviting us to contribute a piece on this topic. This paper was written while DHP was in receipt of a Philip Leverhulme Prize.

244

9780826443595_Ch11_Final_txt_print.indd 244

2/9/2012 9:36:23 PM

12 DAVID HUME AND THE ARGUMENT TO DESIGN Andrew Pyle 1. BACKGROUND This proof always deserves to be mentioned with respect. It is the oldest, the clearest, and the most accordant with the common reason of mankind. It enlivens the study of nature, just as it itself derives its existence and gains ever new vigour from that source. It suggests ends and purposes, where our observation would not have detected them by itself, and extends our knowledge of nature by means of the guiding concept of a special unity, the principle of which is outside nature. This knowledge again reacts on its cause, namely upon the idea which has led to it, and so strengthens the belief in a supreme Author of Nature that the belief acquires the force of an irresistible conviction. (I. Kant, Critique of Pure Reason)1 The argument to design, often misleadingly called ‘the argument from design’, is indeed, as Kant says, distinguished both by its great antiquity and its manifest psychological appeal to minds like ours. From time immemorial, it seems, men have contemplated the beauty and order of the heavens, or admired the intricate functional contrivances of the parts of organisms, and seen such order and contrivance as clear evidence of the workings

of an intelligent cause. Behind the more formal statements of the argument always lurks the same rhetorical question: how could such things have come to be without the operation of an intelligent cause? Once we attempt to spell out this reasoning in more detail, we immediately see that the argument to design is not a single argument but a family of related arguments. The distinguishing feature of such arguments is their strict empiricism. The premises must be independently verifiable matters of empirical fact such as the regular planetary orbits of the Newtonian system, or the adaptation of structure to function in the parts of animals. No atheist will dispute the fact that the planets move in mathematically precise orbits, or that the parts of the human eye are admirably arranged to make vision possible. The inference from these facts is intended to conform to our best canons of inductive reasoning, making the argument to design – so we are told – a truly ‘scientific’ proof of the existence of God. It is this strict adherence to the principles of empiricism that distinguishes the argument to design from most of the other attempted proofs of the existence of God, and explains its widespread appeal to thinkers suspicious of the metaphysical subtleties of the ontological and cosmological proofs.

245

9780826443595_Ch12_Final_txt_print.indd 245

2/9/2012 9:36:30 PM

DAVID HUME AND THE ARGUMENT TO DESIGN Arguments to design, I have claimed, are best thought of as a family of related inferences. Here are three common and familiar ways of setting up the argument: (1) As an inference to best explanation, where the intelligent designer is posited as the best explanation of the facts to be explained (e.g. the functional contrivance of the parts of animals). (2) As an argument from improbability, where the claim is that the facts to be explained (various aspects of the order of the universe) would be staggeringly improbable if no intelligent cause were involved; if our cosmos were simply the result, for example, of the random collisions of atoms. (3) As an argument from analogy, in which the resemblance between organisms and artefacts (which are known to be products of intelligent design) grounds an inference to a similar intelligent cause for the former. Clear statements of what we would recognize as versions of the argument to design can be found throughout the history of human thought. In antiquity, the great physician Galen wrote a lengthy treatise On the Usefulness of the Parts of the Body, noting the manifestations of intelligent design throughout human anatomy.2 In the famous dialogue On the Nature of the Gods by the Roman philosopher and orator Cicero, the Stoic Balbus spells out the ‘infinite improbability’ version of the design argument against his Epicurean opponent Velleius.3 In the Middle Ages, another version of the design argument, from the goal-directedness of the parts of nature, makes an appearance as the fifth of the famous ‘Five Ways’ of St Thomas Aquinas.4 In seventeenth-century Britain, scientists of the calibre of Robert Boyle5 and Isaac Newton6 gave it their seal

of approval, and their work was in turn cited by the Boyle Lecturers.7 In early eighteenthcentury Scotland, the argument was championed by the celebrated Newtonian Colin MacLaurin (1698–1746), whose Account of Sir Isaac Newton’s Philosophical Discoveries (1748) endorses the ‘plain argument’ for the existence of a deity ‘from the evident contrivance and fitness of things for one another’.8 This ‘Newtonian’ theism was very much part of the intellectual background of the young David Hume – one commentator has suggested MacLaurin as a possible model for the character of Cleanthes, the advocate of natural theology in Hume’s Dialogues.9 And of course in our own time, the hypothesis of ‘intelligent design’ is still defended, by a variety of advocates, as a ‘scientific’ rival to Darwin’s theory of evolution by natural selection.

2. READING THE DIALOGUES [Hume] Some of my friends flatter me, that it [the Dialogues] is the best thing I ever wrote. I have hitherto forborne to publish it, because I was of late desirous to live quietly and keep remote from all clamour: For though it be not more exceptionable than some things I had formerly published; yet you know some of these were thought very exceptionable; and in prudence, perhaps, I ought to have suppressed them. I there introduce a sceptic, who is indeed refuted, and at last gives up the argument, nay confesses that he was only amusing himself by all his cavils; yet before he is silenced, he advances several topics, which will give umbrage, and will be deemed very bold and free, as well as much out of the common road.10 [Hume] On revising them [the Dialogues] (which I have not done these fifteen

246

9780826443595_Ch12_Final_txt_print.indd 246

2/9/2012 9:36:30 PM

DAVID HUME AND THE ARGUMENT TO DESIGN years) I find that nothing can be more cautiously and more artfully written.11 Although the Dialogues were published only posthumously, we know from Hume’s correspondence that he intended to publish them during his lifetime. With this in mind, we need to take seriously his statement, in the above-quoted letter to Adam Smith, that the work was ‘cautiously’ and even ‘artfully’ written. Irreligious and anticlerical writers had to use, at this period, a variety of literary devices to get their views into print. One might resort to irony or to misdirection,12 one might report the views of real or imaginary doubters, or of course one might write in dialogue form. The obvious advantage of the dialogue form is that it enables the author to distance himself from the views of any particular character. We know that Hume, who had a deep and abiding love of the classics, used Cicero’s De Natura Deorum as his model in writing the Dialogues.13 If asked for his own opinion, he might have simply repeated Cicero’s own warning: Those who ask for my own opinion on every question merely show excessive curiosity. In a discussion of this kind our interest should be centred not on the weight of the authority but on the weight of the argument. Indeed the authority of those who set out to teach is often an impediment to those who wish to learn.14 Despite such warnings, readers of a work of philosophy written in dialogue form will still, almost inevitably, ask the obvious question: ‘Who speaks for the author?’ Hume himself says, in a well-known letter to his friend Gilbert Elliot, that ‘I make Cleanthes the hero of the dialogue’,15 and of course the Dialogues

close with Philo’s notorious ‘U-turn’ and a summing-up by the narrator Pamphilus in Cleanthes’ favour. Contemporary reviewers were not taken in for one moment by these literary devices – the testimony of Hume’s earliest critics is solidly in favour of the ‘Philo is Hume’ reading of the work.16 Following Dugald Stewart,17 however, a number of commentators came to endorse the rival ‘Cleanthes is Hume’ reading, which in turn was flatly rejected by Norman Kemp Smith in the introduction to his 1935 edition of the Dialogues. Kemp Smith’s own verdict was very clear: I shall contend that Philo, from start to finish, represents Hume; and that Cleanthes can be regarded as Hume’s mouthpiece only in those passages in which he is explicitly agreeing with Philo, or in those other passages in which, while refuting Demea, he is also being used to prepare the way for one or other of Philo’s independent conclusions.18 After Kemp Smith’s attack, the ‘Cleanthes is Hume’ reading of the Dialogues has found few defenders. This interpretation seems deaf to the wit and irony of the work, and blind to the ‘cautious’ and ‘artful’ way in which it was presented. But if we reject the ‘Cleanthes is Hume’ reading, does it follow that we must accept Kemp Smith’s ‘Philo is Hume’ reading? By no means. A number of recent commentators have argued that the Dialogues should be read simply as dialogues, without seeking for a single authorial voice or a simple authorial message.19 Modern opinion is divided between the ‘Philo is Hume’ reading and the ‘No single authorial voice’ reading. But if Philo has by far the best of the arguments, then this may be, in the final analysis, a distinction without a difference. We set out to read the Dialogues simply as dialogues,

247

9780826443595_Ch12_Final_txt_print.indd 247

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN and end up being convinced that Philo was right on most or all of the key points at issue.

3. CLEANTHES’ FIRST STATEMENT OF THE ARGUMENT TO DESIGN [Cleanthes] Look around the world: Contemplate the whole and every part of it: You will find it to be nothing but one great machine, subdivided into an infinite number of lesser machines, which again admit of subdivisions, to a degree beyond what human senses can trace and explain. All these various machines, and even their most minute parts, are adjusted to each other with an accuracy, which ravishes into admiration all men, who have ever contemplated them. The curious adapting of means to ends, throughout all nature, resembles exactly, though it much exceeds, the productions of human contrivance; of human design, thought, wisdom, and intelligence. Since therefore the effects resemble each other, we are led to infer, by all the rules of analogy, that the causes also resemble; and that the Author of nature is somewhat similar to the mind of man; though possessed of much larger faculties, proportioned to the grandeur of the work, which he has executed. By this argument a posteriori, and by this argument alone, do we prove at once the existence of a deity and his similarity to human mind and intelligence. (DNR 2.143) The three characters in the Dialogues, Demea, Cleanthes and Philo are all in at least verbal agreement that there is a God, previously defined as the cause of the existence and order of our universe. Setting aside the literally incredible doubts of the most

extreme sceptics, all men agree both that there is an orderly physical universe, and that there must be some cause of its existence and order.20 So Cleanthes’ first statement of the design argument is intended to prove to Demea and Philo not that ‘there is a God’ (which neither of them ever denies), but that ‘[t]he relation between God and the world is like that between a human engineer and his creations’. It is by this analogical argument, says Cleanthes, that we prove God’s natural attributes such as wisdom and forethought, and (hopefully) His moral attributes such as justice and benevolence. Cleanthes invites us to ‘[l]ook around the world’, and to contemplate both the whole and its parts. The universe as a whole, he says, is ‘one great machine’; its parts are ‘an infinite number of lesser machines’. Two different aspects of order are being invoked here. In describing the physical universe as ‘one great machine’, Cleanthes is presumably referring to the Newtonian model of the solar system. This manifests order in the sense of perfect regularity. The Newtonian achievement of explaining the orbits of planets, moons and even comets in terms of a few simple mathematical laws was one of the most spectacular intellectual feats of all time, and made a deep impression on eighteenth-century thinkers. But the parts of the solar system show no clear indications of being machine-like in the sense of having their parts designed for a purpose. We can contemplate the rings of Saturn, or the moons of Jupiter, and have no notion of what they might be for, and indeed no great assurance that they are for anything at all. If we seek for evidence of intelligent contrivance we must turn our gaze from the whole to the parts, to the ‘lesser machines’ of which the ‘great machine’ is composed. The argument shifts from cosmology to biology.

248

9780826443595_Ch12_Final_txt_print.indd 248

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN That the parts of organisms serve definite functions, and can usefully and fruitfully be regarded as existing for the functions they serve, is one of the central assumptions of the biological sciences.21 Teleological assumptions of this kind have been the staple of biological reasoning from Galen on the kidney to Harvey on the valves in the veins. Exactly as in a well-designed machine, each part or organ has the structure it needs in order to serve its given function. It is here in biology that the argument to design finds its clearest evidence and its strongest grounds. Here we take ourselves to know the ends (survival and of course reproduction) that must be served by the parts of organisms, and are thus in a position to make tolerably well-informed judgements about excellence of design. A cat needs its acute senses, sharp claws, swift reflexes and superb agility in order to be a successful hunter of mice. This ‘curious adapting of means to ends’ in nature, says Cleanthes, is a precise analogue of the similar (though markedly inferior) adaptation of means to ends in human engineering. By the rules of analogy we are entitled to infer that as the known cause of the watch is to the watch, so is the unknown cause of the cat to the cat. Philo responds to Cleanthes’ first statement of the argument to design with two objections. The first is straightforward, and simply illustrates an obvious weakness of arguments by analogy. The strength of an analogical argument, he reminds us, depends on the degree of similarity of the cases. If you exaggerate the resemblances and overlook or play down real differences, you will fall into errors of over-generalization. He illustrates his point with a well-chosen example from the life sciences. Once the circulation of the blood had been established in animals, an analogous ‘circulation of the sap’ was supposed to provide the key to the physiology of

plants. But the analogy was spurious – plants turn out to have striking differences from animals. This particular argument from analogy yielded only errors (DNR 2.144). Philo’s second objection goes deeper. Our inferences from effects to causes are, he insists, always based on our past experience, and on the patterns and regularities we have observed in our previous experience of the world. If we see a house, Cleanthes, we conclude, with the greatest certainty, that it had an architect or builder, because this is precisely that species of effect, which we have experienced to proceed from that species of cause. But surely you will not affirm, that the universe bears such a resemblance to a house, that we can with the same certainty infer a similar cause, or that the analogy is here entire and perfect. The dissimilitude is so striking, that the utmost you can pretend to is a guess, a conjecture, a presumption concerning a similar cause . . . . (DNR 2.144) When Cleanthes protests at Philo’s use of terms like ‘guess’ and ‘conjecture’ to characterize his argument, Philo responds by insisting that his objection is simply the consequence of taking the empiricist theory of knowledge seriously. As far as a priori reasoning goes, says Philo, anything can cause anything: order, arrangement, or the adjustment of final causes is not, in itself, any proof of design; but only so far as it has been experienced to proceed from that principle. For aught we can know a priori, matter may contain the source or spring of order originally, within itself, as well as mind does . . . . (DNR 2.146) To restate his argument in a manner consistent with empiricist principles, Philo explains,

249

9780826443595_Ch12_Final_txt_print.indd 249

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN Cleanthes must claim that experience teaches us that ‘there is an original principle of order in mind, not in matter’ (DNR 2.146). But this new version of the design argument suffers from two glaring defects. It rests on little more than anthropomorphic prejudice, and it rushes far too quickly from the part to the whole. What peculiar privilege has this little agitation of the brain which we call thought, that we must thus make it the model of the whole universe? Our partiality in our own favour does indeed present it on all occasions: But sound philosophy ought carefully to guard against so natural an illusion (DNR 2.148).

product of intelligent design. Analogy would then give us grounds for thinking that intelligence is present in the other cases (organisms) too. Philo might retort that there are rather striking differences between organisms and artefacts, most notably in their manner of coming-to-be, and that the analogical inference from artefacts to organisms still strikes him as an instance of anthropomorphic prejudice.

4. SECOND STATEMENT OF THE ARGUMENT TO DESIGN: THE ‘IRREGULAR’ INFERENCE

Stone, wood, iron, brass, have not, at this time, in this minute globe of earth, an order or arrangement without human art and contrivance: Therefore the universe could not originally attain its order and arrangement, without something similar to human art. But is a part of nature a rule for another part very wide of the former? Is it a rule for the whole? (DNR 2.149)

[Maclaurin] It strikes us like a sensation; and artful reasonings against it may puzzle us, but it is without shaking our belief. No person, for example, that knows the principles of optics and the structure of the eye, can believe that it was formed without skill in that science; or that male and female in animals were not formed for each other, and for continuing the species.22

Since we have no experience whatsoever of the origin of worlds, Philo concludes, we can draw no inferences at all regarding their causes. The application of the design argument to the universe as a whole seems entirely precarious, resting on no empirical basis in human experience. Cleanthes might conceivably retreat at this point, and restrict the design argument to the ‘lesser machines’ of which the ‘great machine’ is composed, i.e. to the evidence of design manifest in the parts of plants and animals. Here we have a large class of instances of apparent contrivance (in artefacts and organisms alike), and certain knowledge that in one subset of these cases (artefacts) the contrivance is the

[Cleanthes] Consider, anatomize the eye: Survey its structure and contrivance; and tell me, from your own feeling, if the idea of a contriver does not immediately flow in upon you with a force like that of sensation. The most obvious conclusion surely is in favour of design; and it requires time, reflection, and study, to summon up those frivolous, though abstruse, objections which can support infidelity. Who can behold the male and female of each species, the correspondence of their parts and instincts, their passions and their whole course of life before and after generation, but must be sensible, that the propagation of the species is intended by nature? (DNR 3.154)

250

9780826443595_Ch12_Final_txt_print.indd 250

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN Part III of the Dialogues begins with Cleanthes regrouping, and marshalling arguments for a counter-attack against Philo’s scepticism. Just as some works of art please us although they break the rules, so too, Cleanthes insists, some arguments will persuade us, even if cast in ‘irregular’ form (DNR 3.155). The similarity between the works of nature and those of art is ‘selfevident and undeniable’, the objections of the sceptics mere quibbles, like those of the ancient Greek philosophers who famously denied the existence of motion. Doubts of this kind are to be met not by counterargument but by the presentation of striking examples. Cleanthes launches Part III with his two famous thought-experiments: the voice from the clouds and the living library. Even without the support of a generalization from experience, Cleanthes insists, we would all hear the voice from the clouds as conveying a message, and the volumes of the living library as an indication of an authorial intelligence. When Cleanthes, here following Maclaurin almost to the letter, claims that the idea of a contriver for the eye strikes us ‘with a force like that of sensation’, he is directing our attention to two aspects of the resulting belief: passivity and immediacy. Certain objects, he thinks, simply strike us as products of design, even without any background of relevant experience or any conscious processes of inductive or analogical reasoning. If this is correct, the argument to design may rest on the secure foundation of a natural belief, and be as such immune to sceptical doubts. Cleanthes, it appears, thinks that the belief in design may be natural to humans. But could this also have been Hume’s opinion? There is no doubt at all that Hume accepts the existence of some natural beliefs. Key examples are our belief in the reality of

the external world and our assumption of the uniformity of nature. Our confidence is entirely unshaken when we come to realize that neither reason nor experience provides adequate grounds for these assumptions. Sceptical arguments against external world realism and the uniformity of nature can puzzle us, but they can never convince us. Why not? Because, as Hume says in a famous line from the Treatise, ‘Nature, by an absolute and uncontrollable necessity has determin’d us to judge as well as to breathe and feel’ (THN 1.4.1.7 / 183).23 If natural judgements are involuntary, they cannot be subject to the usual norms of belief-formation. The sceptic tells me that I ought to suspend judgement regarding the truth or falsehood of the proposition ‘This ball will fall if I release it’, but experience tells me that I do in fact form the strong and confident expectation that it will fall. Natural judgements thus enjoy a privileged position in our network of beliefs, and a special immunity to sceptical doubts. But could Hume have thought that our belief in design is another natural belief, on a par with our beliefs in the external world, the regularity of nature, and the existence of other minds? This ‘natural belief’ interpretation of the Dialogues was first proposed in a famous paper by Ronald Butler,24 and has recently been championed by Stanley Tweyman.25 The ‘natural belief’ interpretation has two significant virtues. It fits precisely with Cleanthes’ strategy in Part III: there can be little doubt that Cleanthes is encouraging us to regard belief in an intelligent designer of our world as so deeply engrained in human nature that it would be folly to attempt to doubt it. On this view, the sceptic’s doubts are akin to Zeno’s doubts about the reality of motion, mere artefacts of the philosopher’s study, not objections that demand serious study and detailed point-by-point rebuttal. And the

251

9780826443595_Ch12_Final_txt_print.indd 251

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN ‘natural belief’ interpretation explains neatly Philo’s famous U-turn in Part XII, where he admits that his earlier doubts were mere ‘cavils’ and that he too accepts the existence of intelligent design in nature. If belief in design is a natural belief, the sceptic can consistently reject the arguments but accept their conclusion. So Philo’s notorious U-turn is no problem for the ‘natural belief’ reading of the Dialogues. But could Hume himself have regarded the belief in an intelligent designer of nature as another natural belief? The evidence, both textual and philosophical, is negative. In the introduction to the Natural History of Religion, Hume notes that theism is a very widespread belief among human societies, but he then adds this crucial qualification:

by this well-known letter against the ‘natural belief’ reading of the Dialogues is very powerful. Hume here admits a ‘propensity’ of the human mind to see design in nature, but adds that:

The belief of invisible, intelligent power has been very generally diffused over the human race, in all places and all ages; but it has neither perhaps been so universal as to admit of no exception, nor has it been, in any degree, uniform in the ideas, which it has suggested. Some nations have been discovered, who entertained no sentiments of Religion, if travellers and historians are to be credited; and no two nations, and scarce any two men, ever agreed precisely in the same sentiments. It would appear, therefore, that this preconception springs not from an original instinct or primary impression of Nature . . . since every instinct of this kind has been found absolutely universal in all nations and ages. . . . The first religious principles must be secondary (NHR 134).

Our tendency to anthropomorphize aspects of nature is a staple of the poetic imagination, but in the natural sciences it has been a fertile source of errors such as the medieval belief that ‘Nature abhors a vacuum’. There may be a natural tendency here, but Hume insists that it is only a natural weakness that can and should be controlled, and could never provide a legitimate ground of assent. The classic study of this question is that of John Gaskin.27 He lists four marks or criteria of a Humean natural belief. They must be:

. . . unless that Propensity towards it [the design argument] were as strong and universal as that to believe in our Senses and Experience, [it] will still, I am afraid, be esteem’d a suspicious Foundation. We must endeavour to prove that this Propensity is somewhat different from our Inclination to find our own Figures in the Clouds, our Face in the Moon, our Passions and Sentiments even in inanimate Matter. Such an inclination may, & ought to be control’d, & can never be a legitimate Ground of Assent.26

(1) Ordinary beliefs of common life. (2) Incapable of rational justification in the face of sceptical doubts. (3) Indispensable in practice for everyday life. (4) Universally held and culturally invariant among humans.

In 1751, while composing the Dialogues, Hume found himself engaged in correspondence on precisely this issue with his friend Gilbert Elliot. The textual evidence provided

Our beliefs in the external world, the uniformity of nature, and the existence of other minds fit all four of these criteria extremely

252

9780826443595_Ch12_Final_txt_print.indd 252

2/9/2012 9:36:31 PM

DAVID HUME AND THE ARGUMENT TO DESIGN well. Theism fares much worse: it manifestly fails criteria (3) and (4), and arguably fails (1) and (2) as well. We should conclude, with Gaskin, that the belief in design cannot reasonably be construed as a ‘natural belief’ in the strong Humean sense. It may be a ‘natural belief’ in some weaker sense, in that we humans do seem to have a strong propensity to read nature in anthropomorphic terms, but this may be a mere natural weakness of significance for anthropology but not for metaphysics or epistemology.

5. THE DESIGN ARGUMENT AND THEISM [Philo] In a word, Cleanthes, a man, who follows your hypothesis, is able, perhaps, to assert, or conjecture, that the universe, sometime, arose from something like design: But beyond that position he cannot ascertain one single circumstance, and is left afterwards to fix every point of his theology, by the utmost licence of fancy and hypothesis. (DNR 5.168–9) [Kant] Now no one, I trust, will be so bold as to profess that he comprehends the relation of the magnitude of the world as he has observed it (alike as regards both extent and content) to omnipotence, of the world order to supreme wisdom, of the world unity to the absolute unity of its Author, etc. Physico-theology is therefore unable to give any determinate concept of the supreme cause of the world, and cannot therefore serve as the foundation of a theology which is itself in turn to form the basis of religion.28 In many respects, Part V of the Dialogues is the most straightforward and unproblematic. Here the sceptical Philo launches a series of

arguments for the conclusion that natural theology is unable to provide the sort of support that the defenders of established religions such as Christianity demanded from it. If we simply start from the phenomena of nature, and argue without bias and preconception, we have no way of establishing the traditional list of attributes of the God of the great monotheist tradition. Indeed, if we take the argument from analogy seriously, it may in fact provide better support for a variety of heterodox views. The first and most obvious problem concerns God’s supposed infinitude. Since we are inferring the cause from the effect, and the effect (at least ‘so far as it falls under our cognisance’) is finite, we ought to infer that the cause is finite (DNR 5.166). Since an infinite mind – if such a thing exists – would be radically unlike ours, the analogy is much stronger and clearer for a finite designer than for an infinite one. Cleanthes does not respond to this objection in Part V, but later in the Dialogues, in the discussion of evil in Part XI, he shows himself quite prepared to drop all talk of infinity in our discussion of the divine attributes, and embrace the hypothesis of a finite god. ‘I have been apt to suspect,’ Cleanthes admits, that ‘the frequent repetition of the word infinite, which we meet with in all theological writers, to savour more of panegyric than of philosophy, and that any purposes of reasoning, and even of religion, would be better served, were we to rest contented with more accurate and more moderate expressions’ (DNR 11.203). But Philo is only warming up. Along with God’s infinity must go his supposed perfection. If we are arguing from the phenomena of nature without bias or prejudice, our world might well strike us as manifestly imperfect in any number of respects. And even if we were to judge the design of

253

9780826443595_Ch12_Final_txt_print.indd 253

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN our world to be perfect (which would fly in the face of experience) we would have no grounds for assuming a corresponding perfection in the designer. In human technology, perfection of design is generally the result of a long, slow process of trial and error, with each generation of designers building on the work of its predecessor. The marks of design in our world might reflect a similar history: ‘Many worlds might have been botched and bungled, throughout an eternity, ’ere this system was struck out’ (DNR 5.167). We have no way, Philo concludes, of assigning ‘where the probability lies’ between the rival hypotheses: excellent first-time design versus a long story of world-building by trial and error. But if we take the argument from analogy seriously, the probability must lie with world-building by trial and error rather than perfect first-time design. The unity of God is obviously a central pillar of the theology of the great monotheist tradition running through Judaism, Christianity and Islam. But, Philo asks Cleanthes, ‘what shadow of an argument . . . can you produce, from your hypothesis, to prove the unity of the Deity?’ (DNR 5.167). In human affairs, many men frequently collaborate to produce great feats of architecture and engineering. If I come across such a work, I might naturally assume that the requisite knowledge and skills were distributed around the design team rather than being concentrated in a single head. Unity of design need not indicate unity of designer. Here we need to note that in his Natural History of Religion, Hume had dismissed polytheism as an ‘arbitrary supposition’, and had endorsed the argument from unity of design to unity of designer (NHR 2.138). So what are we to make of the apparent contradiction between the two texts? If we deny that Philo speaks for Hume, we could see his flirtation with

polytheism as a mere jeu d’esprit, but it is noteworthy that Cleanthes is given no serious reply in defence of monotheism. I think the Philo of the Dialogues is very close to Hume’s own views, in which case the explicit monotheism of the Natural History must be taken with a large pinch of salt. Philo’s final point takes us right into the heart of Christian theology. A man who reasons in accordance with your principles, he warns Cleanthes, has no reason to believe in divine providence: This world, for ought he knows, is very faulty and imperfect, compared to a superior standard; and was only the first rude essay of some infant Deity, who afterwards abandoned it, ashamed of his lame performance; it is the work only of some dependent, inferior Deity; and is the object of derision to his superiors: it is the production of old age and dotage in some superannuated Deity; and ever since his death, has run on at adventures, from the first impulse and active force, which it received from him . . . . (DNR 5.169)

The scholastic philosophers of the Middle Ages drew a distinction between the cause of the coming-to-be of a thing (causa secundum fieri) and the sustaining cause of its ongoing existence (causa secundum esse). The causes of the coming-to-be of a child are its parents, and the cause of the coming-to-be of a building is a builder. But the parents can desert or disown the child, or even die, and the child will continue to exist. By contrast, according to the schoolmen, the cause of the very being of sunlight is the sun: if the sun were to cease to exist, so too would its light. (A modern example might be the relation between a magnet and its magnetic field.) The problem for Cleanthes is that he represents the

254

9780826443595_Ch12_Final_txt_print.indd 254

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN relation between God and his Creation as akin to that between a human designer and his creations, that is, as the relation of causa secundum fieri. But the theologian wants the relation between God and his works to be that of causa secundum esse – in Him, as Saint Paul so memorably puts it, ‘we live and move and have our being’.29 So beneath the surface of this light-hearted gibe about a ‘superannuated deity’, Philo is making a deep point about the inadequacy of the design argument to serve Christian theology. Part V ends with Cleanthes’ response to Philo’s objections. Astonishingly, he claims a sort of victory. Your flights of imagination, he tells Philo, give me pleasure, when I see, that, by the utmost indulgence of your imagination, you never get rid of the hypothesis of design in the universe, but are obliged, at every turn, to have recourse to it. To this concession I adhere steadily; and this I regard as a sufficient foundation for religion. (DNR 5.169) It is the pious Demea who responds with signs of horror, as he sees the implications of this sort of natural theology for religious faith and practice: While we are uncertain, whether there is one Deity or many; whether the Deity or Deities, to whom we owe our existence, be perfect or imperfect, subordinate or supreme, dead or alive; what trust or confidence can we repose in them? What veneration or obedience pay them? To all the purposes of life, the theory of religion becomes altogether useless . . . . (DNR 6.170)

our assistance – he tells us quite clearly that he does not think it helpful or enlightening to insist on the infinitude of the divine attributes. With regard to unity, we must try to find some marks or criteria – within the products of human arts and crafts – which indicate the unity of the designing mind. This seems unpromising, not to say deeply problematic. With regard to perfection, we will find ourselves mired in the problems of evil – of which more anon. But the most fundamental problem for empirical natural theology is Philo’s last and deepest objection. Cleanthes’ argument rests squarely on the analogy between the products of human craft (machines) and the products of supposed divine craft (organisms). But the relation between human craftsmen and their products is always that of causa secundum fieri, and this does not and cannot give theologians the relation between God and man that they want. Even switching from a craft metaphor to a biological one (‘Our Father’) does not help. Our parents will inevitably become old and feeble and eventually die, leaving us to our own resources. As Demea sees clearly, Cleanthes’ style of natural theology is all but useless as a support for orthodox religious belief and practice.

6. ALTERNATIVE POSSIBILITIES

Can we do any better for Cleanthes? With regard to God’s infinity, he does not want

[Philo] A continual circulation of matter in it [the world] produces no disorder: A continual waste in every part is incessantly repaired: The closest sympathy is perceived throughout the entire system: And each part or member, in performing its proper offices, operates both to its own preservation and that of the whole. The world, therefore, is an animal, and the Deity is the soul of the world,

255

9780826443595_Ch12_Final_txt_print.indd 255

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN actuating it, and actuated by it. (DNR 6.170–1) [Philo] But if we must needs fix on some hypothesis; by what rule, pray, ought we to determine our choice? Is there any other rule than the greater similarity of the objects compared? And does not a plant or an animal, which springs from vegetation or generation, bear a stronger resemblance to the world, than does any artificial machine, which arises from reason and design? (DNR 7.177) At the end of Part V, Cleanthes thinks he is winning the argument, since his sceptical opponent Philo never gets rid of ‘the hypothesis of design in the universe’. In Parts VI to VIII, Philo takes his sceptical critique of the argument to design still further, arguing that arguments from analogy can support naturalistic as well as supernatural accounts of the order of our world. The world, says Cleanthes, resembles a machine in various respects, so analogy suggests that its cause is an intelligent designer. But the world, Philo retorts, also resembles an animal or a vegetable, in which case analogy suggests that it owes its origin to generation or vegetation. We favour the hypothesis of intelligent design because it mirrors our human way of designing and making artefacts. But this partiality for the design hypothesis may be mere human prejudice. A race of super-intelligent spiders might favour the hypothesis that the world was spun from the bowels of a vast spider. ‘Why an orderly universe may not be spun from the belly as well as from the brain, it will be difficult for him [Cleanthes] to give a satisfactory reason.’ (DNR 7.180–1) Within our experience, objects showing functional contrivance of parts arise from four distinct sources: Reason (the pocket watch), Instinct (the spider’s web), Generation (the cat) and

Vegetation (the oak tree). To insist that Reason must somehow be involved behind the scenes in all cases in which order arises from Instinct, Generation and Vegetation is merely question-begging. Philo goes on, in Part VIII, to revisit the old Epicurean hypothesis, albeit with a few important modifications. If we assume the existence of a finite stock of atoms, recombining endlessly over infinite periods of time, then all possible combinations of atoms will be tried out sooner or later (DNR 8.182). Among those possible combinations of atoms, some will be relatively stable, able to maintain themselves in their respective environments. If all possible combinations have been tried out over the ages, and only viable ones have survived and reproduced, then, says Philo, we have the beginnings of a reductive and non-teleological account of the apparent teleology of nature: It is in vain, therefore, to insist upon the uses of the parts in animals or vegetables, and their curious adjustment to each other. I would fain know how an animal could subsist, unless its parts were so adjusted? Do we not find, that it immediately perishes whenever this adjustment ceases, and that its matter corrupting tries out some new form? (DNR 8.185) Here Philo is simply repeating Book Five of Lucretius’ De Rerum Natura, that classic statement of Epicurean Atomism.30 When Cleanthes attacks the Epicurean hypothesis, Philo is quick to respond that he is not in fact championing it but merely raising it as an alternative hypothesis to Cleanthes’ theism. The same warning holds, of course, for the world-animal and world-vegetable hypotheses. As a sceptic, says Philo, it is my

256

9780826443595_Ch12_Final_txt_print.indd 256

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN business to come up with alternative scenarios to the claims of the dogmatist. My dogmatic opponent, he says, claims to know that hypothesis h1 is true, on the basis of evidence e. In a purely sceptical spirit, says Philo, I raise the suggestion that the same evidence e is equally compatible with rival hypotheses h2, h3, or h4. For the sceptic, the more rival hypotheses in play the better. And if we are rational, we will have to take the ‘catch-all’ hypothesis (that the truth is not even among the hypotheses we are currently considering) very seriously. As Philo puts it, with perhaps a hint of exaggeration: Without any great effort of thought, I believe that I could, in an instant, propose other systems of cosmogony, which would have some faint appearance of truth; though it is a thousand, a million to one, if either yours or any of mine be the true system (DNR 8.182). Is Philo’s final position one of strict neutrality between the rival hypotheses, leading to a simple suspension of judgement and the conclusion that the whole subject (the origin of worlds) is simply beyond our faculties? Or are there hints and suggestions in the text that would lead us to think that, although all our hypotheses are admittedly precarious, some have better grounds in experience than others? In his recent study Spectres of False Divinity, Thomas Holden argues that Philo (who he takes to be Hume’s spokesman throughout the Dialogues) accepts only what he calls a ‘liminal’ natural theology, in which we know no more of the intrinsic nature of God (defined in relational terms as the cause of our world) than of any unknown object X.31 On this view, Philo’s scepticism leads him to eschew entirely what Holden calls ‘core’ natural theology, that is, attempts

to say anything positive about the specific nature and distinguishing properties of God. This is, of course, a perfectly natural reading of much of what Philo says throughout the Dialogues, and is perhaps also what Philo should say, assuming as Holden does that he is Hume’s spokesman, arguing from Hume’s own epistemological principles. There remain, however, clear hints and suggestions throughout the Dialogues of an anti-theistic agenda that goes beyond the mere sceptical suspension of judgement. In Part II Philo betrays a clear materialist bias in describing thought as ‘this little agitation of the brain’ (DNR 2.148). In Part IV he suggests that we should cease our search for causes with the material world (DNR 4.161–2), rather than seeking to explain its order in terms of the order of a supposedly prior ideal world (the divine mind). In Part VI he argues that the great advantage of the world-animal and world-vegetable hypotheses is that they rid us of the notion, ‘repugnant to common experience’, of a mind existing without a body (DNR 6.171). In Part VII we are told that the evidence of experience is that Reason (intelligence) often arises from Generation (sex), but never vice versa (DNR 7.179). In all these cases, the naturalistic counter-hypothesis to Cleanthes’ theism is raised within the context of a sceptical agenda, allowing Philo plenty of room for evasion and retreat. If challenged, he can say that he is not positively advocating these naturalistic and materialistic hypotheses – he is merely proposing them as a corrective to the widespread and systematic prejudice in favour of theism, and hence as an aid to the suspension of judgement he is advocating. But the arguments can easily take on a life of their own, and provide a cumulative case for a naturalistic metaphysics. We can readily grant that we do not know the nature

257

9780826443595_Ch12_Final_txt_print.indd 257

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN and properties of the cause or causes of our world. We can grant that all of our hypotheses have very low probabilities given the evidence. But Philo does seem to be claiming that, if we view the evidence of experience without bias or preconception, naturalistic hypotheses will seem more plausible (better grounded in experience) than theistic ones. This comparative judgement (the probability of h2 given e is greater than the probability of h1 given e) is of course perfectly compatible with the assignment of very low probabilities to both of the rival hypotheses. On this view, Philo can be regarded as a sceptic with a definite leaning towards naturalism and atheism, rather than a pure sceptic who is entirely neutral between the rival hypotheses.

7. THE INFERENCE PROBLEM OF EVIL [Philo] Look around this universe. What an immense profusion of beings, animated and organised, sensible and active! You admire this prodigious variety and fecundity. But inspect a little more narrowly these living existences, the only beings worth regarding. How hostile and destructive to each other! How insufficient all of them for their own happiness! How contemptible or odious to the spectator! The whole presents nothing but the idea of a blind nature, impregnated by a great vivifying principle, and pouring forth from her lap, without discernment or parental care, her maimed and abortive children (DNR 11.211). Parts X and XI of the Dialogues deal with the issue of moral and physical evil, and follow thematically from Part V, in which Philo has raised the problem for Cleanthes of our inability to infer the traditional attributes

of God from the argument to design. If Cleanthes’ type of natural theology is to serve a religion such as Christianity, it ought to provide evidence from experience for God’s moral attributes such as benevolence and justice. A God without benevolence (who did not intend the happiness of his creatures), or a God without justice (who did not reward the virtuous and punish the vicious) could hardly be a suitable object of worship for us. So if experience does not provide evidence of benevolence and justice built into the very fabric of the natural world, Cleanthes has a problem. Philosophical discussions of the so-called ‘problem of evil’ often focus on the compatibility or otherwise of the various evils (moral and physical) of our world with the assumption of an all-powerful, all-knowing, and benevolent God. But this is only a ‘problem’ for those people already committed to a particular variety of theism. The problem facing a natural theologian such as Cleanthes is the quite distinct inference problem. Examine the natural and human worlds without bias or preconception, and ask what moral properties, if any, one can infer about its designer and creator. In the discussion of evil in the Dialogues, Philo raises the consistency problem, but the heart of his disagreement with Cleanthes concerns – quite properly – the inference problem. This is the real problem facing the empirical natural theologian. In Section 11 of the first Enquiry, Hume had introduced a ‘friend who loves sceptical paradoxes’ (EHU 11.1 / 132), and who wondered whether experience testifies to a God or gods motivated by justice. Here in the Dialogues, the focus is on the question of whether the phenomena of the natural world give us reasons for inferring benevolence in its designer and creator. The pious Demea and the sceptical Philo agree that the sufferings of men and

258

9780826443595_Ch12_Final_txt_print.indd 258

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN beasts outweigh their pleasures in this world. Demea thinks that this world is merely a porch or antechamber to the next, and that our sufferings in this world are merely trials, to be amply compensated for in the afterlife (DNR 10.199). Cleanthes, to his credit, sees that this stratagem will simply not do: we are, after all, trying to infer the moral properties of God (the gods) from experience, not to fit the facts of experience to an arbitrarily chosen hypothesis (DNR 10.199–200). But, says Philo, taking our world at face value, you would never infer benevolence in its creator: I will allow, that pain or misery in man is compatible with infinite power and goodness in the Deity, even in your sense of these attributes: What are you advanced by all these concessions? A mere possible compatibility is not enough. You must prove these pure, unmixed, and uncontrollable attributes from the present mixed and confused phenomena, and from these alone. A hopeful undertaking! (DNR 10.201) Philo goes on to list a variety of natural phenomena that seem, at least on the face of it, difficult to reconcile with the hypothesis of a God or gods motivated by benevolence. At this point the Manichean theory of two warring gods, one good and the other evil, presents itself as a serious possibility.32 This hypothesis, says Philo, has ‘more probability than the common hypothesis’, but seems inconsistent with the apparent unity of design manifest in our world. The more natural conclusion is that God (the gods) is (are) indifferent to human well-being: There may four hypotheses be framed concerning the first causes of the universe: that they are endowed with perfect

goodness, that they have perfect malice, that they are opposite and have both goodness and malice, that they have neither goodness nor malice. Mixed phenomena can never prove the two former unmixed principles. And the uniformity and steadiness of general laws seems to oppose the third. The fourth, therefore, seems by far the most probable (DNR 11.212).

How seriously are we meant to take this argument? Remember that it is Philo who is speaking here, and Philo’s overall position is sceptical, arguing throughout the Dialogues that we lack the data to establish any system of cosmology. Is he here abandoning his sceptical principles? Thomas Holden thinks that Philo cannot intend this argument to be taken at face value, and that it is best read as a ‘parody’ of Cleanthes.33 Holden’s judgement is in part determined by his own distinction between ‘core’ and ‘liminal’ natural theology, and his insistence that Philo/Hume utterly rejects the former. If we regard Holden’s distinction as un-Humean, we can allow Philo to advance hypotheses in the domain of ‘core’ natural theology, albeit with the usual sceptical disclaimers. Philo’s message, as best we can read it, seems to be the following. If we were perfectly wise, we would realize that theology is simply beyond our powers, and abandon the subject altogether. But we are not perfectly wise, and will always find ourselves speculating about what, if anything, lies behind the world of experience. If forced to assess the relative merits of competing hypotheses about God (defined as ‘the cause of the existence and/or order of the universe’), what else can we do other than to ask which of our rival hypotheses seems best to fit with the course of experience? On these grounds, we find ourselves favouring the

259

9780826443595_Ch12_Final_txt_print.indd 259

2/9/2012 9:36:32 PM

DAVID HUME AND THE ARGUMENT TO DESIGN hypothesis of God as the world-soul against the hypothesis of an immaterial God separate from the world, because experience testifies clearly against the hypothesis of disembodied minds. Such a thing has no analogy within the world of experience. Similarly, Philo can say, our experience of the mixture of pleasures and pains that life presents seems most easily reconciled with the hypothesis that God (the gods) is (are) indifferent to our weal or woe. There is simply less argumentative work to be done to reconcile this hypothesis with experience than there is for the others. The chief weakness in Holden’s account is his failure even to address the clear – and very marked – difference in tone between Philo’s discussion of the inference from nature to God’s intelligence and the corresponding inference to God’s benevolence. In our past disputes, says Philo to Cleanthes, I had to exercise my ‘sceptical and metaphysical subtilty’, and even to invent ‘mere cavils and sophisms’ to elude the force of the inference from the adaptations of the parts of animals and vegetables to the existence of an intelligent designer (DNR 10.202–3). A neutral observer might have concluded that Cleanthes was right, and that experience does present strong reasons for belief in intelligent design in nature. On the issue of God’s moral properties, however, the sceptic triumphs, and the would-be natural theologian is defeated:

your philosophical subtilties against the dictates of plain reason and experience (DNR 10.202). I conclude that we have no reason to accept Holden’s characterization of Philo’s argument for the moral indifference of God (the gods) as mere parody. There are of course gaps and weaknesses in the argument, obvious shortcuts and overlooked objections. But Philo is always giving us two judgements, not one. There is the over-arching sceptical agenda, with its clear message that we just do not know enough to make confident judgements in matters of theology. But there is also the equally clear insistence that, if we are forced to rank a variety of competing hypotheses, we can do so only on the basis of the testimony of experience. And unless we want to ‘tug the labouring oar’, we must admit that ‘the gods don’t care about us’ has better grounds in experience than ‘the gods love us’.

8. CONCLUSION

[Philo] But there is no view of human life, or of the condition of mankind, from which, without the greatest violence, we can infer the moral attributes, or learn that infinite benevolence, conjoined with infinite power and infinite wisdom, which we must discover by the eyes of Faith alone. It is your turn now to tug the labouring oar, and to support

[Philo] If the whole of Natural Theology, as some people seem to maintain, resolves itself into one simple, though somewhat ambiguous, at least undefined proposition, that the cause or causes of order in the universe probably bear some remote analogy to human intelligence: If this proposition be not capable of extension, variation, or more particular explication. If it afford no inference that affects human life, or can be the source of any action or forbearance: And if the analogy, imperfect as it is, can be carried no farther than to the human intelligence; and cannot be transferred, with any appearance of probability, to the other qualities of the mind: If this really be the

260

9780826443595_Ch12_Final_txt_print.indd 260

2/9/2012 9:36:33 PM

DAVID HUME AND THE ARGUMENT TO DESIGN case, what can the most inquisitive, contemplative, and religious man do more than to give a plain, philosophical assent to the proposition, as often as it occurs; and believe that the arguments, on which it is established, exceed the objections which lie against it? (DNR 12.227) In the final part of the Dialogues, Philo performs his notorious ‘U-turn’ and tells Cleanthes that, for all his stated doubts and objections, he too has a firm sense of the evidence of divine intelligence provided by ‘the inexplicable contrivance and artifice of nature’ (DNR 12.214). Galen’s inference from the functional complexity of the parts of animals to the existence of an intelligent designer is, Philo now tells us, all but irresistible, and only strengthened by subsequent advances in anatomy and physiology (DNR 12.215). My doubts and objections, he tells Cleanthes, were prompted by ‘my love of singular arguments’, and were never intended to be taken seriously (DNR 12.214). What are we to make of Philo’s U-turn? If the ‘natural belief’ interpretation of the Dialogues were correct, it would resolve the problem and dispel any air of mystery about Philo’s (and Hume’s?) position. On this reading, Philo demolishes the arguments for intelligent design but holds fast to the belief in intelligent design, because that belief is grounded in human nature rather than in rational argument. (Likewise, of course, Hume persuades us that we have no rational grounds for believing in the uniformity of nature, but is emphatically not trying to persuade us to become inductive sceptics.) The U-turn, on this reading, would be simply the reaffirmation of the natural belief in the face of sceptical doubts that are dismissed as literally incredible. Instead of meeting the sceptic’s arguments, we just look at the functional

contrivance of the parts of the eye and allow the resulting belief in design to strike us ‘with a force like that of sensation’ (DNR 3.154). But we have already seen both textual and philosophical reasons to reject the ‘natural belief’ interpretation. So how else are we to explain Philo’s U-turn? One reading that stays close to the text is to take Philo’s words as perfectly sincere, and to see Philo and Cleanthes as converging, in Part 12 of the Dialogues, on a position that might best be described as a sort of ‘weak deism’. On this reading, some of Philo’s doubts are, as he himself says, mere ‘cavils and sophisms’, not intended to furnish serious objections to the argument to design. If Philo’s doubts were never 100 per cent serious, then talk of a ‘U-turn’ is exaggerated. Our overall subjective probability for the proposition that ‘Nature manifests intelligent design’ is somewhat reduced when we survey Philo’s ingenious list of alternative possibilities, but remains high enough to count as firm assent. This reading has been championed by John Gaskin in his important study, Hume’s Philosophy of Religion.34 It fits with the text not only of the Dialogues, but also of the Natural History of Religion, where the argument to design is given in plain unvarnished form. ‘The whole frame of nature,’ we are there told, ‘bespeaks an intelligent author; and no rational enquirer can, after serious reflection, suspend his belief a moment with regard to the primary principles of genuine Theism and Religion.’ (NHR 134) Although Gaskin’s interpretation remains perfectly permissible as a reading of the text – as no doubt Hume intended – it does overlook some of the textual evidence, and in particular fails to do justice to the ‘artful’ way in which the Dialogues were written. When Philo concludes that ‘the cause or causes of order in the universe probably bear

261

9780826443595_Ch12_Final_txt_print.indd 261

2/9/2012 9:36:33 PM

DAVID HUME AND THE ARGUMENT TO DESIGN some remote analogy to human intelligence’ (DNR 12.227), we naturally think that he is making a significant concession to Cleanthes. But Philo has already told us that there is ‘a certain degree of analogy among all the operations of nature’, including ‘the rotting of a turnip, the generation of an animal, and the structure of human thought’ (DNR 12.218). So the final tentative conclusion of a probable case for a ‘remote’ analogy between the cause of the world and human intelligence is in fact no concession at all. The ‘U-turn’ is exaggerated not (as Gaskin thinks) because Philo’s initial doubts were never serious, but because his eventual conversion is merely verbal. Two other lines of evidence support this mildly ‘atheistic’ reading of the Dialogues. In presenting his case against Cleanthes’ theism, Philo advances and defends a variety of naturalistic hypotheses regarding the cause of order in the natural world. Officially, these naturalistic hypotheses are put forward within a sceptical agenda, as alternative possibilities intended not to elicit our assent but to induce us to suspend judgement. But the arguments can easily take on a life of their own, and provide a cumulative case for the conclusion that some form of naturalism is more probable – on the evidence – than theism.35 Contemporaries of Hume such as Joseph Priestley were not fooled by Philo’s supposed U-turn. In his Examination of Mr Hume’s Dialogues (1780) he sums up Hume’s intentions as atheistic: although Philo . . . advances nothing but common-place objections against the belief of a God, and hackneyed declamation against the plan of Providence, his antagonists are seldom represented as making any satisfactory reply. And when, at the last, evidently to save appearances,

he relinquishes the argument, on which he had expatiated with so much triumph, it is without alleging any sufficient reason; so that the arguments are left, as no doubt the writer intended, to have their full effect in the mind of the reader. And although the debate seemingly closes in favour of the theist, the victory is clearly on the side of the atheist.36

The final piece of evidence against Philo’s supposed theism comes from a key passage from the end of Part IV of the Dialogues.37 The argument to design is presented as an inference to best explanation, giving an account of the material order of nature in terms of the ideal order of the contents of the divine mind. But, says Philo, ideal order requires explanation every bit as much as material order (DNR 4.160). Neither Reason nor Experience provides any grounds for our prejudice in favour of ideal order over material. If we demand an explanation for functional complexity as such, we must demand such an explanation for the contents of the divine mind, which will launch us into a regress. But if this is the case, why not stop with the material world rather than searching for the cause of its order in a supposedly prior ideal world? ‘An ideal system, arranged of itself, without a precedent design, is not a whit more explicable than a material one, which attains its order in like manner; nor is there any more difficulty in the latter supposition than in the former’ (DNR 4.164). Since Cleanthes gives only a feeble response, this unanswered objection to the theistic hypothesis is left – as no doubt Hume intended – to influence our eventual understanding of the Dialogues as a whole. Whether the Dialogues teach a weak deism, or agnosticism,38 or, as I have argued, show evidence of clear leanings towards

262

9780826443595_Ch12_Final_txt_print.indd 262

2/9/2012 9:36:33 PM

DAVID HUME AND THE ARGUMENT TO DESIGN naturalism and atheism, one point remains completely clear. The argument to design, whatever its persuasive power, affords ‘no inference that affects human life, or can be the source of any action or forbearance’ (DNR 12.227). This often-overlooked qualification to the conclusion is in many respects the key to the whole work. The all-important conclusion that Hume clearly wants us to take from his critical examination of the argument to design is that the argument gives no support whatsoever to any established religion, and can provide no shred of a basis for any religious claims about our moral duties and obligations. As Thomas Holden forcefully argues in his impressive new book, Hume was definitely a moral atheist,39 believing that morality was a purely human affair, grounded in our human sentiments and affections, and inapplicable for deep reasons of principle to the cause of the existence and order of the universe.

7

8

9

10

NOTES

11

1

12

2

3

4

5

6

I. Kant, Critique of Pure Reason, trans. N. Kemp Smith (London: Macmillan, 1976), p. 520. Galen, On the Usefulness of the Parts of the Body, trans. Margaret Talladge May (New York: Cornell University Press, 1968). Cicero, On the Nature of the Gods, trans. H.C.P. McGregor, intro. J.M. Ross (Harmondsworth: Penguin, 1972), p. 161. Cicero’s work was of course Hume’s model for his own Dialogues. For the ‘Five Ways’, see Aquinas, Summa Theologiae, Questions on God, ed. B. Davies and B. Leftow (Cambridge: Cambridge University Press, 2006), ‘Does God Exist?’, pp. 20–7. See, for example, R. Boyle’s Usefulness of Natural Philosophy (Works, 6 vols, ed. T. Birch, London, 1772), vol. 5, p. 427. I. Newton, Mathematical Principles of Natural Philosophy, trans. Motte, revised F. Cajori,

13

14 15

16

17

(Berkeley, Los Angeles and London: University of California Press, 1934), 2 vols, General Scholium, vol. 2, p. 544: ‘This most beautiful system of the sun, planets, and comets, could only proceed from the counsel and dominion of an intelligent being’. For the argument from the functional contrivance of the parts of animals, see Opticks (New York, Dover, 1952), Query 28, pp. 369–70: ‘How come the Bodies of Animals to be contrived with so much Art, and for what ends were their several Parts? Was the Eye contrived without Skill in Opticks, or the Ear without Knowledge of Sounds?’ For the place of the design argument in the Boyle Lectures, see my Introduction to a new reprint of G. Burnet’s A Defence of Natural and Revealed Religion, being an Abridgement of the Sermons Preached at the Lecture Founded by Robert Boyle, 4 vols (Bristol: Thoemmes, 2000), vol. 1, pp. xxvii–xxx. C. MacLaurin, An Account of Sir Isaac Newton’s Philosophical Discoveries [1748] (New York and London: Johnson Reprint Corporation, 1968), p. 381. See R.H. Hurlbutt III, Hume, Newton, and the Design Argument (Lincoln, NB: University of Nebraska Press, 1965). Hume to William Strahan (his publisher), 8 June 1776 (LDH 2.322, 525). Hume to Adam Smith, 15 August 1776 (LDH 2.334, 538). An obvious example would be Voltaire dedicating his play about the prophet Mohammed to the Pope. See C. Battersby, ‘The Dialogues as Original Imitation: Cicero and the Nature of Hume’s Scepticism’, in D.F. Norton, N. Capaldi and W. Robison (eds), McGill Hume Studies (San Diego: Austin Hill Press, 1979), pp. 239–52. Cicero, On the Nature of the Gods, p. 73n3. Hume to Gilbert Elliot, 10 March 1751 (LDH 1.154n10, 72). For a brief summary of this evidence, see my Hume’s Dialogues concerning Natural Religion: A Reader’s Guide (London: Continuum, 2006), pp 135–6. For the original reviews, see S. Tweyman (ed.), Hume on Natural Religion (Bristol: Thoemmes, 1996). D. Stewart, Collected Works, ed. Sir W. Hamilton, 11 vols (Edinburgh: Constable, 1854–60), vol. 1, p. 605.

263

9780826443595_Ch12_Final_txt_print.indd 263

2/9/2012 9:36:33 PM

DAVID HUME AND THE ARGUMENT TO DESIGN 18

19

20

21

22

23

24

25

26 27

N. Kemp Smith, introduction to Hume’s Dialogues Concerning Natural Religion (Indianapolis: Bobbs-Merrill, 1947), p. 59. See, for example, W.L. Sessions, Reading Hume’s Dialogues: A Veneration for True Religion (Bloomington and Indianapolis: Indiana University Press, 2002), p. 212, and Michel Malherbe, ‘Hume and the Art of Dialogue’, in M.A. Stewart and J.P. Wright (eds), Hume and Hume’s Connexions (Philadelphia: Pennsylvania University Press, 1995), pp. 201–23. In the Treatise, of course, external world scepticism is dismissed in the following famous passage. ‘We may well ask, What causes induce us to believe in the existence of body? But ’tis in vain to ask, Whether there be body or not? That is a point, which we must take for granted in all our reasonings.’ (THN 1.4.2.1 / 187; ‘Of Scepticism with Regard to the Senses’). Kant claims that this is a regulative assumption of the life sciences. We cannot, he says, prove that it is correct, but biologists must proceed as if it were known to be correct. See the Critique of Judgment, trans. W.S. Pluhar (Indianapolis: Hackett, 1987), pp. 280–3. Maclaurin, An Account of Sir Isaac Newton’s Philosophical Discoveries; cited by Hurlbutt, Hume, Newton, and the Design Argument, p. 42. THN 1.4.1.7 / 183, ‘Of Scepticism with Regard to Reason’. R.J. Butler, ‘Natural Belief and the Enigma of Hume’, Archiv für die Geschichte der Philosophie 42 (1960), pp. 73–100. S. Tweyman, Scepticism and Belief in Hume’s Dialogues concerning Natural Religion (Dordrecht: Kluwer, 1986), chap. 8, pp. 121–56. LDH 1.155, 72. J.C.A. Gaskin, Hume’s Philosophy of Religion, 2nd edn (London: Macmillan, 1988), p. 109.

28 29 30

31

32

33

34

35

36

37

38

39

Kant, Critique of Pure Reason, p. 523. Acts 17: 28. Titus Lucretius Carus, On the Nature of Things, trans. Sir R. Melville, with an Introduction by D. and P. Fowler (Oxford: Oxford University Press, 1999), Bk 5, lines 837–48. T. Holden, Spectres of False Divinity: Hume’s Moral Atheism (Oxford: Oxford University Press, 2010), chap. 2, pp. 19–47. According to Pierre Bayle in his famous Dictionnaire historique et critique, the consistency problem of evil cannot be solved by philosophy or rational theology. (See especially the articles ‘Manicheans’ and ‘Paulicans’.) It is only the eye of Faith that can reconcile the manifest evils of our world with the belief in a perfectly good God. Whether Bayle’s fideism was sincere or merely tactical continues to divide scholarly opinion. T. Holden, Spectres of Divinity, p. 169n30. For more traditional readings, see Gaskin, Hume’s Philosophy of Religion, p 72n27, and D. O’Connor, ‘Scepticism and Hume’s Atheistic Preference’, Hume Studies 29 (2003), pp. 267–82. Gaskin, Hume’s Philosophy of Religion, chap. 7, pp 120–31. I present a sketch of this case in my Reader’s Guide to Hume’s Dialogues (London: Continuum, 2006), pp. 130–2. Joseph Priestley, Letters to a Philosophical Unbeliever, quoted from Tweyman (ed.), Hume on Natural Religion, p. 81. Philo’s argument is repeated – without any acknowledgment of Hume – in Richard Dawkins’s The Blind Watchmaker (London: Penguin, 1988), p. 141. J. Noxon, ‘Hume’s Agnosticism’, The Philosophical Review 73 (1964), pp. 248–61. Holden, Spectres of False Divinity, pp. 4–9.

264

9780826443595_Ch12_Final_txt_print.indd 264

2/9/2012 9:36:33 PM

13 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF David O’Connor

In his ‘Introduction’ to The Natural History of Religion, Hume tells us that two questions about religion ‘challenge our attention’(NHR 134). One is about religion’s ‘foundation in reason’, the other is about its ‘origin in human nature’. Taken together, Hume’s investigations of the two questions make a strong case against the reasonableness of religious belief. Hume, however, does not claim this as his aim in either investigation, as the outcome of either investigation, or as the aim or outcome of the two together, and there are passages in his work that, taken at face value, suggest the opposite, namely, that the reasonableness of religious belief is not in question for him at all. For instance, taking up the first question in Dialogues concerning Natural Religion, Hume frames it as a conversation just about the nature of God, not the existence of God, and has the sceptical Philo mouthing agreement with the pious Demea that, to ‘reasonable men’, the existence of God is ‘unquestionable and self-evident’ and proved by the cosmological argument (DNR 2.142). And when he refers to that first question in The Natural History, Hume’s words are that no reasonable person could doubt the evidence of design in nature or the perfection of

the designer. But the idea that the existence of God is not at issue in Hume’s discussion of design in nature or of the nature of the supposed designer is undercut by his actual investigation of those things in the Dialogues and in Section XI of An Enquiry concerning Human Understanding. The result is that Hume’s praise of the design argument is mere praise1 and Philo’s professed agreement with Demea comes to a good deal less than it seems to promise, reaching at best to what J.C.A. Gaskin calls an ‘attenuated deism’, something far